Download - TER WEI SHEAN University of Malaya
THE EFFECT OF VOCAL WARM-UP ON VOICE QUALITY OF UNTRAINED FEMALE SINGERS IN MALAYSIA
TER WEI SHEAN
CULTURAL CENTRE
UNIVERSITY OF MALAYA KUALA LUMPUR
2020Univ
ersity
of M
alaya
THE EFFECT OF VOCAL WARM-UP ON VOICE QUALITY OF UNTRAINED FEMALE SINGERS IN
MALAYSIA
TER WEI SHEAN
THESIS SUBMITTED IN FULFILMENT OF THE
REQUIREMENTS FOR THE DEGREE OF MASTER OF PERFORMING ARTS (MUSIC)
CULTURAL CENTRE UNIVERSITY OF MALAYA
KUALA LUMPUR
2020 Univers
ity of
Mala
ya
ii
UNIVERSITY OF MALAYA
ORIGINAL LITERARY WORK DECLARATION
Name of Candidate :
Matric No :
Name of Degree :
Title of Thesis :
Field of Study :
Ter Wei Shean
ROA180012
Master of Performing Arts (Music)
The Effect of Vocal Warm-up on Voice Quality of
Untrained Female Singers in Malaysia
Music Education
I do solemnly and sincerely declare that:
(1) I am the sole author/writer of this Work; (2) This Work is original; (3) Any use of any work in which copyright exists was done by way of fair dealing and for permitted purposes and any excerpt or extract from, or reference to or reproduction of any copyrighted work has been disclosed expressly and sufficiently and the title of the Work and its authorship have been acknowledged in this Work; (4) I do not have any actual knowledge nor do I ought reasonably to know that the making of this work constitutes an infringement of any copyrighted work; (5) I hereby assign all and every right in the copyright to this Work to the University of Malaya (“UM”), who henceforth shall be the owner of the copyright in this Work and that any reproduction or use in any form or by any means whatsoever is prohibited without the written consent of UM having been first had and obtained; (6) I am fully aware that if in the course of making this Work I have infringed any copyright whether intentionally or otherwise, I may be subject to legal action or any other action as may be determined by UM.
Candidate’s Signature Date:
Subscribed and solemnly declared before,
Witness’s Signature Date:
Name:
Designation:
Univers
ity of
Mala
ya
iii
THE EFFECT OF VOCAL WARM-UP ON VOICE QUALITY OF UNTRAINED
FEMALE SINGERS IN MALAYSIA
ABSTRACT
Professional singers and singing teachers regard vocal warm-up as vital. However,
very few studies have evaluated this effect quantitatively. This study aimed to evaluate
the effect of the vocal warm-up on the singing voice quality through acoustic parameters
of Jitter, Shimmer and harmonics-to-noise ratio (HNR) and to assess whether scientific
evidence supports the warm-up procedure, whether an untrained singer can benefit from
singing vocal warm-up exercises in a similar way as a trained singer and whether the
vocal warm-up should be encouraged or bypassed. 40 untrained female singers were
recorded twice while uttering the vowel /a/, /o/ and /i/ in two different pitches: Low- A3
(220.0 Hz) and High- C5 (523.2 Hz) for at least five seconds. The recordings were
collected before and after a 20 minutes vocal warm-up session. Results showed significant
variations in the average values of the parameters measured. A decrease was detected in
comparison with the average value of Jitter and Shimmer before and after the vocal warm-
up, whereas HNR increased. The results of this study provide valid support for the
advantageous effect of vocal warm-up on the voice quality of untrained female singers
and present acoustic analysis as a valuable and sensitive tool for quantifying this effect.
The positive effects of the findings indicated that the vocal warm-up should be
encouraged and not bypassed. As for the vowel effect, per cent jitter, per cent shimmer
and harmonics-to-noise ratio (HNR) may not be useful for the description of acoustic
differences between vowels.
Keywords: vocal warm-up, voice quality, acoustic parameters
Univers
ity of
Mala
ya
iv
KEBERKESANAN PEMANASAN SUARA TERHADAP KUALITI VOKAL
PENYANYI TIDAK TERLATIH DI MALAYSIA
ABSTRAK
Pemanasan vokal adalah sangat penting bagi penyanyi profesional and guru vokal
untuk mengekalkan kesihatan suara mereka. Namun, masih terdapat kurang kajian yang
dijalankan secara kuantitatif dalam bidang ini. Kajian ini bertujuan untuk menyelidik
keberkesanan pemanasan suara terhadap kualiti vokal dengan menggunakan parameter
akustik Jitter, Shimmer and harmonics-to-noise ratio (HNR). Selain itu, kajian ini
dijalankan untuk menilai sama ada bukti saintifik menyokong prosedur pemanasan vokal,
menilai sama ada penyanyi yang tidak terlatih boleh mendapat manfaat daripada
menyanyi latihan pemanasan vokal dengan cara yang sama seperti penyanyi terlatih dan
menilai sama ada pemanasan vokal harus digalakkan atau dilangkau. 40 penyanyi wanita
yang tidak terlatih direkodkan dua kali sambil mengucapkan vokal /a/, /o/ dan /i/ dalam
dua nada: Nada rendah-A3 (220.0 Hz) dan Nada tinggi-C5 (523.2 Hz) untuk sekurang-
kurangnya lima saat. Rakaman telah dikumpulkan sebelum dan selepas sesi pemanasan
vokal selama 20 minit. Keputusan menunjukkan variasi ketara dalam nilai purata
parameter akustik yang diukur. Penurunan nilai purata Jitter dan Shimmer dikesan dengan
perbandingan sebelum dan selepas pemanasan suara, sedangkan HNR meningkat. Hasil
kajian ini memberikan sokongan yang sah mengenai kebaikan dan kepentingan
pemanasan vokal terhadap kualiti suara penyanyi wanita yang tidak terlatih dan
membuktikan analisis akustik sebagai alat yang berharga dan sensitif untuk mengukur
kesan ini. Kesan positif penemuan dalam kajian ini menunjukkan bahawa pemanasan
suara harus digalakkan dan tidak dilangkau. Keputusan kajian ini mencadangkan peratus
Jitter dan Shimmer serta harmonics-to-noise ratio (HNR) mungkin tidak berguna untuk
perihalan perbezaan akustik antara vokal.
Kata Kunci: pemanasan vokal, kualiti vokal, parameter akustik
Univers
ity of
Mala
ya
v
ACKNOWLEDGEMENTS
First and foremost, I would like to express my sincere gratitude to my supervisor
Dr. Wong Kwan Yie for her guidance, encouragement and patience throughout my
master study. She has been there providing her kind support and guidance at all times
and has given me invaluable guidance, inspiration and suggestions in my quest for
knowledge. She has given me all the freedom to pursue my research, while silently
and non-obtrusively ensuring that I stay on course and do not deviate from the core of
my research.
Next, I would like to express my deepest appreciation to all the volunteers who
were chosen to participate in this research study. The experiment could not have been
successfully conducted without their passionate participation and contribution of time.
I extend special thanks to my Mom and my dear friend, Madam Lee Yin Kim for their
assistance and support in conducting the experiment. I would like to thank my friends
who have supported me in numerous ways. Last but not least, I am deeply grateful to
my family, especially my parents, for their unconditional love, constant support
encouragement, never-ending patience, without whose love, this work would never be
possible.
Univers
ity of
Mala
ya
vi
TABLE OF CONTENTS
Abstract ............................................................................................................................ iii
Abstrak ............................................................................................................................. iv
Acknowledgements ........................................................................................................... v
Table of Contents ............................................................................................................. vi
List of Figures .................................................................................................................. ix
List of Tables..................................................................................................................... x
List of Symbols and Abbreviations ................................................................................. xii
List of Appendices ......................................................................................................... xiii
CHAPTER 1: INTRODUCTION .................................................................................. 1
1.1 Background of the study .......................................................................................... 1
1.2 Statement of the Problem......................................................................................... 5
1.3 Research Objectives................................................................................................. 7
1.4 Purpose of the Study ................................................................................................ 8
1.5 Significance of the Study ......................................................................................... 9
1.6 Definition of Terms ............................................................................................... 10
1.7 Conceptual Framework .......................................................................................... 12
1.8 Research Questions ................................................................................................ 13
1.9 Delimitations of the Study ..................................................................................... 14
1.10 Organization of the Study ...................................................................................... 15
CHAPTER 2: LITERATURE REVIEW .................................................................... 16
2.1 Introduction............................................................................................................ 16
2.2 Production of Voice ............................................................................................... 18
2.3 Laryngeal Biomechanics ....................................................................................... 20
2.4 Perturbation Measures – Jitter, Shimmer & Harmonics-to-noise ratio (HNR) ..... 25
2.5 Acoustic Voice Analysis Using Praat Software .................................................... 31
Univers
ity of
Mala
ya
vii
2.6 Physiologic Effects of Vocal Warm-up ................................................................. 32
2.7 Vocal Tract Tuning, Resonance, and Acoustics .................................................... 36
2.8 The Singer’s Formant ............................................................................................ 40
2.9 Measurement of the Singer’s Formant .................................................................. 43
2.10 Disordered Voices ................................................................................................. 45
2.11 Vocal warm-up ...................................................................................................... 47
2.11.1 Traditional Vocal Warm-up & Semi-Occluded Vocal Tract Warm-Up .. 49
2.12 Summary……………………………………………………………………... ….54
CHAPTER 3: RESEARCH METHODOLOGY ....................................................... 56
3.1 Research Design .................................................................................................... 56
3.2 Subjects .................................................................................................................. 56
3.3 Recording procedure and instrumentation ............................................................. 57
3.4 Acoustic Analysis .................................................................................................. 59
3.5 Statistical Analysis................................................................................................. 59
3.6 Pilot Study ............................................................................................................. 60
3.7 Reliability ............................................................................................................... 60
3.8 Data Analysis Procedure........................................................................................ 61
3.8.1 Research Question One ............................................................................ 61
3.8.2 Research Question Two ............................................................................ 61
3.8.3 Research Question Three .......................................................................... 62
3.8.4 Research Question Four ........................................................................... 62
3.8.5 Research Question Five ............................................................................ 63
CHAPTER 4: RESULTS .............................................................................................. 64
4.1 Research Question One.......................................................................................... 67
4.2 Research Question Two ......................................................................................... 69
4.3 Research Question Three ....................................................................................... 71
Univers
ity of
Mala
ya
viii
4.4 Research Question Four ......................................................................................... 73
4.4.1 Jitter .......................................................................................................... 73
4.4.2 Shimmer ................................................................................................... 76
4.4.3 Harmonics-to-noise ratio .......................................................................... 79
4.5 Research Question Five ......................................................................................... 82
4.5.1 Jitter .......................................................................................................... 82
4.5.2 Shimmer ................................................................................................... 85
4.5.3 Harmonics-to-noise ratio .......................................................................... 88
CHAPTER 5: DISCUSSIONS, CONCLUSIONS AND RECOMMENDATIONS 91
5.1 Summary ................................................................................................................ 91
5.2 Discussions of the Research Questions ................................................................. 92
5.2.1 Research Question One ............................................................................ 92
5.2.2 Research Question Two ............................................................................ 93
5.2.3 Research Question Three .......................................................................... 94
5.5.4 Research Question Four ........................................................................... 95
5.5.5 Research Question Five ............................................................................ 96
5.3 Implications of the Study ....................................................................................... 98
5.4 Conclusions ........................................................................................................... 99
5.5 Recommendations for Future Research ............................................................... 102
REFERENCES ............................................................................................................ 103
APPENDIX .................................................................................................................. 117
Appendix A - Participant Questionnaire ....................................................................... 117
Appendix B – Vocal Warm-up Exercises ..................................................................... 118
Appendix C – Statistical Analysis Calculated by SPSS ............................................... 119
Univers
ity of
Mala
ya
ix
LIST OF FIGURES
Figure 1 : Conceptual Framework................................................................................ 12
Figure 2 : One group Pretest - Posttest design (Sage, 2019) ........................................ 56
Figure 3 : Example of Vocal Acoustic Analysis by Praat ............................................ 65
Figure 4 : Jitter, Shimmer and HNR parameters determined using Praat .................... 66
Figure 5 : Group means values and standard deviation for the Jitter parameter at Low pitch level (A3) , pre and post vocal warm-up ............................................ 75
Figure 6 : Group means values and standard deviation for the Shimmer parameter at Low pitch level (A3), pre and post vocal warm-up .................................... 78
Figure 7 : Group means values and standard deviation for the Harmonics-to-noise ratio (HNR) parameter at Low pitch level (A3), pre and post vocal warm-up ... 81
Figure 8 : Group means values and standard deviation for the Jitter parameter at High pitch level (C5), pre and post vocal warm-up ............................................. 84
Figure 9 : Group means values and standard deviation for the Shimmer parameter at High pitch level (C5), pre and post vocal warm-up .................................... 87 Figure 10 : Group means values and standard deviation for the Harmonics-to-noise ratio (HNR) parameter at High pitch level (C5), pre and post vocal warm-up .... 90
Univers
ity of
Mala
ya
x
LIST OF TABLES
Table 1: Mean Values and Standard Deviations (in Parentheses) of Jitter parameter for all the Vowels /a/, /o/ and /i/ in the Two Pitch-Conditions, before and after Vocal Warm-Up ......................................................................................................................................... 67
Table 2: Paired T-test Results of the Jitter parameter for all the Vowels /a/, /o/ and /i/ in the Two Pitch-Conditions, before and after Vocal Warm-Up ........................................ 67
Table 3: Mean Values and Standard Deviations (in Parentheses) of Shimmer parameter for all the Vowels /a/, /o/ and /i/ in the Two Pitch-Conditions, before and after Vocal Warm-Up......................................................................................................................... 69
Table 4: Paired T-test Results of the Shimmer parameter for all the Vowels /a/, /o/ and /i/ in the Two Pitch-Conditions, before and after Vocal Warm-Up .................................... 69
Table 5: Mean Values and Standard Deviations (in Parentheses) of Harmonics-to-noise ratio (HNR) parameter for all the Vowels /a/, /o/ and /i/ in the Two Pitch-Conditions, before and after Vocal Warm-Up .................................................................................... 71
Table 6: Paired T-test Results of the Harmonics-to-noise ratio (HNR) parameter for all the Vowels /a/, /o/ and /i/ in the Two Pitch-Conditions, before and after Vocal Warm-Up ......................................................................................................................................... 71
Table 7: Mean Values and Standard Deviations (in Parentheses) of Jitter for the Vowels /a/, /o/ and /i/ in the Low Pitch-Condition, before and after Vocal Warm-Up ............... 73
Table 8: Paired T-test Results of the Jitter parameter classified by vowels.................... 73
Table 9: Mean Values and Standard Deviations (in Parentheses) of Shimmer for the Vowels /a/, /o/ and /i/ in the Low Pitch-Condition, before and after Vocal Warm-Up .. 76
Table 10: Paired T-test Results of the Shimmer parameter classified by vowels ........... 76
Table 11: Mean Values and Standard Deviations (in Parentheses) of Harmonics-to-noise ratio for the Vowels /a/, /o/ and /i/ in the Low Pitch-Condition, before and after Vocal Warm-Up......................................................................................................................... 79
Table 12: Paired T-test Results of the Shimmer parameter classified by vowels ........... 79
Table 13: Mean Values and Standard Deviations (in Parentheses) of Jitter for the Vowels /a/, /o/ and /i/ in the High Pitch-Condition, before and after Vocal Warm-Up ............... 82
Table 14: Paired T-test Results of the Jitter parameter classified by vowels.................. 82
Table 15: Mean Values and Standard Deviations (in Parentheses) of Shimmer for the Vowels /a/, /o/ and /i/ in the High Pitch-Condition, before and after Vocal Warm-Up .. 85
Univers
ity of
Mala
ya
xi
Table 16: Paired T-test Results of the Shimmer parameter classified by vowels ........... 85
Table 17: Mean Values and Standard Deviations (in Parentheses) of Harmonics-to-noise ratio (HNR) for the Vowels /a/, /o/ and /i/ in the High Pitch-Condition, before and after Vocal Warm-Up .............................................................................................................. 88
Table 18: Paired T-test Results of the Harmonics-to-noise ratio (HNR) parameter classified by vowels ........................................................................................................ 88
Univers
ity of
Mala
ya
xii
LIST OF SYMBOLS AND ABBREVIATIONS
HNR : Harmonics-to-noise ratio
PTP : Phonation Threshold Pressure
SPSS : Statistical Package for the Social Sciences
SPR : Singing Power Ratio
TA : Paired Thyroarytenoid
PCA : Posterior Cricoarytenoid
LCA : Lateral Cricoarytenoid
IA : Interarytenoid
CT : Cricothyroid
MHC : Myosin Heavy Chain
ATP : Adenosine Triphosphate
STF : Slow Tonic Muscle Fibres
VFE : Dr. Joseph Stemple’s Vocal Function Exercises
MANOVA : Multivariate Analysis of Variance
ANOVA : Analysis of Variance
MPFR : Maximum Phonation Frequency Range
PPE : Perceived Phonatory Effort
MRI : Magnetic Resonance Imaging
LPC : Linear Predictive Code
F1 : First Formant
F2 : Second Formant
F3 : Third Formant
RAP : Relative Average Perturbation
L3 : Third Formant
Univers
ity of
Mala
ya
xiii
LIST OF APPENDICES
Appendix A: Participant Questionnaire ........................................................................ 117
Appendix B: Vocal Warm-up Exercises ....................................................................... 118
Appendix C: Statistical Analysis Calculated by SPSS ................................................. 119
Univers
ity of
Mala
ya
1
CHAPTER 1
INTRODUCTION
1.1 Background of the study
The human voice is a result of motor activity, which is determined by genetic,
social, cultural, and linguistic factors. The production involves complicated
aerodynamics and biomechanism that requires the three anatomical systems, namely
respiratory, laryngeal, and surgical systems, be neurophysiologically coordinated. The
respiratory system, together with the larynx, generates the subglottal air pressures
necessary to initiate and maintain voicing (Morrison & Rammage, 1994; Titze, 1994;
Verdolini, 1998). The laryngeal system is the vocalisation source, and the supralaryngeal
system modulates the source-generated energy into an acoustic output perceived by the
listener. Speaking and singing utilised these systems; however, singing voice requires the
maximisation of vocal output by increasing respiration, phonation, and resonance
coordination (Morrison & Rammage, 1994; Titze, 1994; Verdolini, 1998). Good singing
may seem an effortless act, but most singers engage in training since singing is an athletic
activity that requires a good condition and a coordinated interaction of various physical
functions (Sataloff, 2015).
Singing voices vary from speaking voices in a variety of acoustic, physical, and
perceptual ways, and many researchers have quantified these differences. The singing
voice has three major units: the lungs, which act as a power supply; the vocal folds, which
is an oscillator; and the vocal tract, which acts as a resonator (Sundberg, 1977; Zhang,
2016). These elements work together to create the voice that is heard in both speech and
singing, but it is the manipulation of the vocal instrument that occurs in singing that sets
the singing and speaking voices apart, for example, during singing, singers learn to
Univers
ity of
Mala
ya
2
execute a series of resonatory and phonatory adjustments that non-singers don’t learn
(McCrea & Morris, 2007). According to Sundberg (1977), the shape and length of the
vocal folds determine the frequencies created by the vocal folds including the amount of
air pressure moving through them. There are muscles stretching and thinning the vocal
folds. When this happens, the frequency will be higher and when they are in the opposite
shape, the frequency is lower. Singers develop control over their frequency range, which
allows them to sing a larger range or several octaves of frequencies and vocal warm-up
allows them to practice the precision of those vocal instrument adjustments (Sundberg,
1977).
Voice pedagogues, voice therapists, singers and scientists alike devote
tremendous time and energy to designing and validating vocal exercises to enhance the
voice quality of a performer. In voice pedagogy, the term “voice quality” refers to the
distinctive characteristics which characterize the singing voice. In an evaluative context,
the same word is often used to denote to what degree a specific vocal production meets
professional quality expectations. Acoustic measures are instrumental in describing voice
qualities (Teixeira & Fernandes, 2014; Sascha & Pascal, 2019).
Standard acoustic measures that are widely used to measure vocal quality are
jitter, shimmer, and Harmonics-to-noise ratio (HNR) (Bausar, Bohlender, Mehta, 2018;
Teixeira & Fernandes, 2014; Sascha & Pascal, 2019). For this study, the independent
variable was vocal warm-up instructions. The acoustic parameters of jitter, shimmer and
harmonics-to-noise ratio (HNR) were the dependent variables. In the description of
normal and dysphonic speakers, the acoustic parameters, jitter and shimmer, have proven
to be useful in measuring the production of sustained vowels, vocal characteristics related
to roughness and hoarseness respectively. These perturbation measures were used to track
and record frequency variability and instability in the signal. The measure of fundamental
Univers
ity of
Mala
ya
3
frequency cycle-to-cycle variation which is known as the vocal perturbation is referred to
Jitter (De Felippe, Grillo & Grechi, 2006; Teixeira, Oliveira, & Lopes, 2013). Jitter
measures were studied by the following authors: Brown et al. (1990) studied sustained
vowel in female subjects; Sabol et al. (1995) collected values for sustained vowels, and
Brown, Rothman, and Sapienza (2000) studied values for sung and spoken vowel.
Shimmer, also known as amplitude perturbation, quantifies the cycle-to-cycle
variability in waveform amplitude (Teixeira, Oliveira, & Lopes, 2013; James & John,
2007). Shimmer measures were studied by the following authors: Brown et al. (2000)
studied sustained vowel /i/ from speaking and singing tasks between male and female
singers and non-singer groups, and Amir, Amir, and Michaeli (2005) investigated
sustained vowels in classical voices. The perturbation measures provide valuable voice
and vocal health information. Jitter (roughness) has been reported mainly due to
insufficient control of vocal fold vibrations. Patients with vocal fold pathologies have a
higher percentage of reported jitter. With reduced vocal fold lesions and glottal resistance,
Shimmer improves. The presence of noise at emission and breathiness is associated with
Shimmer. The hoarseness and breathiness of voice normally execute vocal quality
changes due to deterioration of voice and laryngeal diseases (Teixeira, Oliveira, & Lopes,
2013; James & John, 2007; Mezzedimi et al., 2018).
The noise-related measure which is the harmonics-to-noise ratio (HNR), captures
the relative contributions of all periodic to aperiodic energy in the speech signal.
Harmonics-to noise ratios (HNR) is an objective and quantitative evaluation of the degree
of hoarseness (De Felippe, Grillo & Grechi, 2006; Teixeira, Oliveira, & Lopes, 2013).
Shimmer measures were studied by the following authors: Brown et al. (2000) studied
sustained vowel /i/ from speaking and singing tasks between male and female singers and
non-singer groups, and Amir, Amir, and Michaeli (2005) investigated sustained vowels
Univers
ity of
Mala
ya
4
in classical voices. Normal voices have low additive noise in the voice and characterized
by a high HNR (Yumoto & Gould, 1982; Teixeira, Oliveira, & Lopes, 2013). Greater
harmonic energy or greater signal in the voice represents more effective vocal fold
vibration and hence better voice quality. Thus, increased noise energy within the signal
suggests abnormal vocal function. Harmonics-to-noise ratio (HNR) measures the number
of voice signals and a higher measure of HNR which is associated with decreased noise
energy represents better vocal quality (Christian, 2007; Teixeira, Oliveira, & Lopes, 2013;
Yumoto & Gould, 1982). Reduced measurements of perturbation associated with reduced
variability in the signal and, therefore, better vocal quality.
There is very little information as to how vocal warm-up affects the acoustics
parameter of the singing voice although the vocal warm-up is widespread and
comprehensive in the singing community (Amir, Amir, & Michaeli, 2005). The goal of
the present study was to use acoustic measures to evaluate the effect of vocal warm-up
on the voice quality of untrained female singers in Malaysia.
Univers
ity of
Mala
ya
5
1.2 Statement of the Problem
The computerized acoustic analysis was presented in previous years as a reliable
and successful tool to assess subtle changes in voice quality and stability. The potential
benefit of this tool is to reliably record, calculate and quantify minor variations that are
otherwise difficult to identify (Teixeira & Fernandes, 2014; Sascha & Pascal, 2019).
Standard acoustic measures that are widely used to measure vocal quality are jitter,
shimmer, and Harmonics-to-noise ratio (HNR) (Bausar, Bohlender, Mehta, 2018;
Teixeira & Fernandes, 2014; Sascha & Pascal, 2019). However, conflicting evidence
exists that these acoustic measures can thoroughly evaluate vocal characteristics.
Although the changes in acoustic values were statistically significant, it did not always
indicate clinically significant improvements in vocal performance. For example, the
singing voice has lower shimmer and noise-to-harmonic ratio than the speaking voice,
according to Lundy et al. (2000). However, no significant differences were found in a
different study using these voice quality parameters between singers and untrained
speakers. From the literature, therefore, it is unclear if such acoustic parameters will
contribute to the vocal quality of untrained female singers after the intervention of vocal
warm-up.
Besides that, a major drawback of earlier research is that the vocal warm-up
procedures used differently between the studies. Some research permitted participants to
utilize a personalized warm-up regimen (Amir, Amir & Michaeli, 2005) while others
utilized structured warm-up procedures (Motel, Fisher, & Leydon, 2002; Elliot, Sundberg
& Gramming, 1995). In previous studies, the duration of warm-up procedures differed
greatly between 7 and 30 minutes (Amir, Amir & Michaeli, 2005; Elliot, Sundberg &
Gramming, 1995; Motel, Fisher, & Leydon, 2002). This lack of consistency may lead to
conflicting results between studies. Thus, this research aimed to assess the impact of vocal
Univers
ity of
Mala
ya
6
warm-up on voice quality of untrained female singers using acoustic measures in two-
pitch conditions, A3 (chest register) and C5 (head register) before and after vocal
warmup. A standardized warm-up protocol with a duration of 20-minutes was carried out
in this study. Hypothetically, one can assume that per cent jitter and shimmer will
decrease while harmonics-to-noise ratio (HNR) will increase following a vocal warm-up.
Univers
ity of
Mala
ya
7
1.3 Research Objectives
1. To evaluate the effect of vocal warm-up on the voice quality of untrained female
singers through the acoustic parameter – Jitter.
2. To evaluate the effect of vocal warm-up on the voice quality of untrained female
singers through the acoustic parameter – Shimmer.
3. To evaluate the effect of vocal warm-up on the voice quality of untrained female
singers through the acoustic parameter – Harmonics-to-noise ratio (HNR).
4. To study whether acoustic parameters (Jitter, Shimmer and Harmonics-to-noise
ratio (HNR)) differ with vowel type (/a/, /o/, /i/) at different pitch level – Low
predetermined tone (frequencies) - A3 (220.0 Hz).
5. To study whether acoustic parameters (Jitter, Shimmer and Harmonics-to-noise
ratio (HNR)) differ with vowel type (/a/, /o/, /i/) at different pitch level – High
predetermined tone (frequencies) - C5 (523.2 Hz).
Univers
ity of
Mala
ya
8
1.4 Purpose of the Study
The goal of the study was (1) to establish whether vocal warm-up immediately
affects the vocal quality of untrained female singers using acoustic analysis, both before
and after vocal warm-up in two pitch conditions, A3 (chest register) and C5 (head
register), (2) to assess whether scientific evidence supports the warm-up procedure, and
(3) whether it should be encouraged or bypassed. Although it is well recognised and
practised by professional, warm-ups are sometimes bypassed and neglected by singers
and voice teachers (Amir, Amir & Michaeli, 2005). The null hypothesis of this
investigation is that there will be no difference in vocal quality between the experimental
conditions as measured by dependent variables of acoustic parameters of Jitter, Shimmer,
and Harmonics-to-noise ratio (HNR). The alternative hypothesis is that there will be
differences in these acoustic measurements as a function of vocal warm-up.
Univers
ity of
Mala
ya
9
1.5 Significance of the Study
Singers are called “vocal athletes” because the demands placed on the singer’s
voice are immense. Singers are different from non-singers as singers have higher stamina,
strength and phonatory agility. This enables them to perform complex laryngeal
manoeuvers to fulfil their vocal demands during singing, according to Zeitels, Hillman,
Desloge, Mauri & Doyle (2002). The benefit of singing vocal warm-ups has long been
documented, and untrained singers could apply the techniques used by trained singers to
enhance their singing voices. Nevertheless, minimal empirical evidence is available to
validate the effect of vocal warm-up on the singing voice quality of untrained female
singers in Malaysia. Most of the research focused heavily on trained singers (Amir, Amir,
and Michaeli, 2005; Thorpe et al., 2001; Griffin, Woo, Colton, Casper & Brewer, 1995).
Therefore, there is a need to gain a greater understanding of what is happening to the
voice quantitatively as a result of warm-up and which parameters are affected in the
voices of untrained singers.
This research seeks to determine whether an untrained singer can benefit from
singing vocal warm-up exercises in a similar way as a trained singer, perhaps proving that
anyone can improve their singing with some training. A strong foundation of supportive
research about vocal warm-ups could assist in establishing the need for vocal training and
education of solo voice educators. Understanding the pedagogical results of vocal warm-
ups could benefit voice teachers in implementing best practice in rehearsals.
Univers
ity of
Mala
ya
10
1.6 Definition of Terms
For the present study, the following terms are operationally defined as follows:
Voice Quality
The distinctive characteristics which characterize the singing voice. It involves both
phonatory and resonatory characteristics (McCrea & Morris, 2007). Some of the
descriptions of voice quality are roughness, breathiness, hoarseness and nasality
(Teixeira, Oliveira, & Lopes, 2013; James & John, 2007). Acoustic measures are
instrumental in describing voice qualities (Teixeira & Fernandes, 2014; Lathadevi &
Suresh, 2008). Acoustic analysis is a part of computerized voice laboratories and it is
useful in supplementing the assessment of voice quality and speech analysis (Sascha &
Pascal, 2019).
Jitter
Jitter is a measure of fundamental frequency cycle-to-cycle variations. It is also known as
the vocal perturbation. A decrease in Jitter indicates that there is an improvement in vocal
roughness (Teixeira & Fernandes, 2014; Lathadevi & Suresh, 2008).
Shimmer
Also known as amplitude perturbation, quantifies the cycle-to-cycle variability in
waveform amplitude. A decrease in Shimmer indicates that there is an improvement in
vocal breathiness (Teixeira & Fernandes, 2014; Lathadevi & Suresh, 2008).
Univers
ity of
Mala
ya
11
Harmonics-to-noise ratio (HNR)
An assessment of the ratio that captures the relative contributions of all periodic to
aperiodic energy in the speech signal. An increase in HNR indicates that there is an
improvement in vocal hoarseness (Teixeira & Fernandes, 2014; Lathadevi & Suresh,
2008).
Untrained Singers
Individuals who enjoy singing as a hobby but never had any previous vocal training.
Two pitch-conditions
Low-A3 (220.0 Hz) and High-C5 (523.2 Hz). The low pitch was produced in the chest
register and the high note in the head register (Lee, Kwon, Choi, Lee, Lee & Jin, 2008).
Univers
ity of
Mala
ya
12
1.7 Conceptual Framework
Figure 1: Conceptual Framework
The Effect of Vocal Warm-up on Voice Quality of Untrained Female Singers in
Malaysia
Post-test taken individually
Voice Quality
Dependent Variable
Voice Quality
Dependent Variable
Intervention of Instruction
Vocal Warm-Up
Independent Variable
Independent Variable
Participants (N=40)
Experimental Group
Pre-test taken individually
Voice Quality
Dependent Variable
Dependent Variable
Univers
ity of
Mala
ya
13
1.8 Research Questions:
The following research questions guided this investigation:
1. Does vocal warm-up significantly alter the acoustic parameter - Jitter of the
untrained singing voice?
2. Does vocal warm-up significantly alter the acoustic parameter - Shimmer of the
untrained singing voice?
3. Does vocal warm-up significantly alter the acoustic parameter - Harmonics-to-
noise ratio (HNR) of the untrained singing voice?
4. Do acoustic parameters (Jitter, Shimmer and Harmonics-to-noise ratio (HNR))
differ with vowel type (/a/, /o/, /i/) at different pitch level – Low predetermined
tone (frequencies) - A3 (220.0 Hz)?
5. Do acoustic parameters (Jitter, Shimmer and Harmonics-to-noise ratio (HNR))
differ with vowel type (/a/, /o/, /i/) at different pitch level – High predetermined
tone (frequencies) - C5 (523.2 Hz)?
Univers
ity of
Mala
ya
14
1.9 Delimitations of the Study
While a rigorous design was used in this study, there were still limitations. Caution
should be taken when making generalizations based on the research results presented,
partly because of the following possible factors:
1. The study was only open to untrained female singers.
2. A small sample size of the present study.
3. Only fixed six vocal warm-up exercises for this study.
4. Only fixed three vowels for this study.
5. A control group was not used in this study.
Univers
ity of
Mala
ya
15
1.10 Organization of the Study
In summary, the research questions in this study concentrate on the effect of vocal
warm-up on the voice quality of untrained female singers. This research is structured as
follows, to find answers to the interrelated central questions. Chapter one introduces the
background of the study on vocal warm-up and the acoustic measures of the singing voice.
Chapter two elaborates on laryngeal biomechanics; laryngeal biomechanics; acoustic
analysis of voice - perturbation measures (Jitter, Shimmer and Harmonics-to-noise ratio
(HNR)); acoustic voice analysis using Praat software; the physiological effects of vocal
warm-up; vocal tract tuning, resonance, and acoustics; singer’s formant; measurement of
the singer’s formant; disordered voices; and the effectiveness of vocal warm-up was
completed indicated by the literature review. Chapter three elaborates on the research
method and the data collection procedures. Chapter four elaborates on the results and
analysis after the data is calculated with Praat software and the Statistical Package for the
Social Sciences (SPSS) Version 25. Chapter five concludes with a summary, conclusion
and recommendations for future research.
Univers
ity of
Mala
ya
16
CHAPTER 2
LITERATURE REVIEW
2.1 Introduction
The literature review documented fundamental discussions of voice science and
vocal pedagogy to provide untrained singers with anatomical knowledge of the voice and
the mechanics of vocal mechanics (Ronald, Thomas & Frederick, 2011 Chhetri, Neubauer
& Berry, 2014; Palaparthi et al., 2019; Poletto, Verdun, Stominger & Ludlow, 2004). This
review benefits untrained vocal singers and vocal coaches who seek ways to teach vocal
techniques, as this provides a unique and concise way to answer questions about
fundamental issues of vocal pedagogy, and understand the benefits of singing a vocal
warm-up through the acoustic analysis.
This chapter begins with a review of the studies regarding the production of voice
and laryngeal biomechanics which relate large muscle performance and recovery
processes to that of the intrinsic muscles of the larynx (Chhetri, Neubauer & Berry, 2014;
Palaparthi et al., 2019; Poletto, Verdun, Stominger & Ludlow, 2004). If one is to make
comparisons between the large muscles of the body and the small, intrinsic muscles of
the larynx, background information regarding laryngeal muscle physiology is necessary.
This studies presented in this section focus on the muscle fibre types of the human
thyroarytenoid muscle (TA). Of interest is the fatigability of this primary vocal fold
adductor. Vocal fatigue or voice disorder is a phenomenon reported by both amateur and
professional voice users alike. Vocal pedagogy touts warming up the voice as a means to
improve vocal performance as well as reduce injury and fatigue. Studies on the
effectiveness of vocal warm-up are presented, using perturbation measures to analyze
voice quality in the singing voice.
Univers
ity of
Mala
ya
17
A comprehensive literature review has been done on the voice production;
laryngeal biomechanics; laryngeal biomechanics; acoustic analysis of voice - perturbation
measures (Jitter, Shimmer and Harmonics-to-noise ratio (HNR)); acoustic voice analysis
using Praat software; the physiological effects of vocal warm-up; vocal tract tuning,
resonance, and acoustics; singer’s formant; measurement of the singer’s formant;
disordered voices; and the effectiveness of vocal warm-up was completed. The literature
involved normal voices, trained singing voices, and disordered voices as well as vocal
warm-up exercises.
In the published literature some variations between the voices of singers and non-
singers were identified. Differences have been reported in jitter, shimmer and harmonics-
to-noise ratio (HNR) (Milenkovic, 1987; Brown, Rothman, and Sapienza, 2000),
phonation threshold pressure (Elliot, Sundberg, & Gramming, 1995; McHenry et al.,
2009), and singing power ratio (SPR) (Amir, Amir, and Michaeli, 2005).
Univers
ity of
Mala
ya
18
2.2 Production of Voice
Phonation is the production of voiced sound through vocal fold vibrations. The
sound waves production is a pressure disturbance distributed across time and space
(Zhang, 2016). The larynx is a hollow, tubular structure connected to the top of the trachea
which consists of the intrinsic laryngeal muscles, membranes, cartilages, and ligaments
(Ronald, Thomas & Frederick, 2011). Vocal folds are tissue folds in the throat that are
key to creating sounds through vocalization. They comprised of twin folds of mucous
membrane extending horizontally from back to front across the larynx (Maria, 2007). The
opening space between the vocal folds is referred to the glottis. Vocal fold abduction
involves the movement of a vocal fold away from the glottis due to muscular contraction
within the larynx. Posterior cricoarytenoid (PCA), paired thyroarytenoid (TA), and lateral
cricoarytenoid (LCA), inter arytenoid (IA), and cricothyroid (CT) form part of the
intrinsic laryngeal muscles (Chhetri, Neubauer & Berry, 2014; Palaparthi et al., 2019;
Poletto, Verdun, Stominger & Ludlow, 2004). The intrinsic laryngeal muscles control the
fine motor actions of the larynx. The muscles of the paired thyroarytenoid (TA) form the
majority of the vocal fold body. The paired thyroarytenoid (TA) muscles shorten, thicken
and stiffen the vocal folds upon contraction. Contraction of the CT muscles causes the
vocal folds to lengthen. Contraction of LCA muscles causes vocal fold adduction. The
contraction of the PCA muscles abducts the vocal folds and lowers resistance. The muscle
of the IA (interarytenoid) holds the vocal cords in a closed position after they are joined
together by the LCA muscles. It closes the glottis and seals the posterior glottis (Chhetri,
Neubauer & Berry, 2014; Palaparthi et al., 2019; Poletto, Verdun, Stominger & Ludlow,
2004). Vocal fold vibration is caused by air pressure below the adducted vocal folds
(Sataloff, 2017). The vocal folds sustain self-oscillation and the oscillation produces a
sound (Titze, 1988). For the oscillation of the vocal fold, the activation of intrinsic
Univers
ity of
Mala
ya
19
laryngeal muscles is important as both small and large oscillations of the amplitude need
approximation of the vocal folds. Vocal morphology and the stress distribution in the
tissue are affected by intrinsic laryngeal muscle activation (Palaparthi et al., 2019).
Production of the sound in singing involves a proper positioning of the larynx and
a shaping of the vocal tract with special attention to the placement of the articulators. The
tongue is to remain free of tension to allow for the vocal tract to remain as open as possible
to create an “open throat” perception. By shaping the tip of the tongue to the bottom of
the lower front teeth, a separation of the hyoid bone from the top of the larynx is
accomplished, again aiding in the low larynx position to optimize vocal tract resonance
(DeCaro & Donato, 2006). Another advantage of a low larynx position was that the
distance of the intraglottal tissue decreases as the distance between the hyoid bone and
thyroid cartilage increases, which means that the vocal folds can vibrate more freely, the
mechanical stress on the vocal folds is lessened, and therefore voice output is enhanced
by expanding the laryngeal ventricle and enhancing the energy of the singer’s formant
(Vurma & Ross, 2003). The biomechanics of the larynx should, therefore, be understood
to explain vocal productions and how they are affected by the warm-up.
Univers
ity of
Mala
ya
20
2.3 Laryngeal Biomechanics
It is essential to understand how the laryngeal muscles can be affected by a vocal
warm-up. This section aims to address the role of the laryngeal muscles in healthy vocal
production and how vocal quality can be enhanced.
Biomechanics concepts can be applied to much of the movement of the laryngeal
muscles, which consist of vocal folds that function mechanically (Cooper, Patridge, &
Alipour-Haghighi, 1993). The laryngeal biomechanics is a complex interaction of solid
mechanics, fluid mechanics, and acoustics (Titze, 2000). Solid mechanics is the study of
rigid bodies, while fluid mechanics is the study of the movement of liquids and gases.
Solids, liquids, and gases are made of tiny particles, and the bulk movement of a large
particle system can be described by mechanical principles. The aggregated particles may
deform, translate, or rotate (Titze, 2000). When the intrinsic, extrinsic, and supplemental
muscles contract, the active forces in the larynx arise, which are then spread across the
vocal fold tissues (Titze, 2000).
Paired thyroarytenoid (TA) muscles and cricothyroid (CT) muscles are the most
prominent intrinsic laryngeal muscles. They are the primary muscles regulating vocal
pitch. Lengthening the vocal folds is mainly achieved by using the CT muscles. It moves
the thyroid cartilage forward and thereby stretches the vocal folds, increases stress, and
raises the vocal pitch. The TA resides within the vocal folds, which makes them shorter
and stiffer and has a nonlinear effect on the vocal pitch. The body layer of the vocal folds
is formed by TA muscles, while CT muscles control the tension in the vocal fold
membranes (Deguchi et al., 2011; Poletto, Verdun, Stominger & Ludlow, 2004). The TA
muscle is responsible for the muscle actions during a speech while the CT muscle plays
with higher fundamental frequencies, for example, singing in the head register and
transition of registers (Watson, 2009; Poletto, Verdun, Stominger & Ludlow, 2004).
Univers
ity of
Mala
ya
21
The vocal warm-up can be measured in numerous ways, for example, what will
happen to the vocal fold tissues after warm-up. According to McHenry, Johnson & Foshea
(2009), the vocal warm-up is a submaximal exercise that is known to have similar warm-
up effects for physical activity. Physical activity warm-up is designed to reduce the risk
of injury, improve skeletal muscle dynamics, and increase athlete performance (Woods,
Bishop & Jones, 2007). The possibilities of physiological improvements during physical
exercise warm-up include: (1) Increased skeletal muscle speed, strength, and relaxation
through faster metabolism; (2) Reduced internal viscosity of active skeletal muscles led
to increased movement efficiency and smoother contractions; (3) Facilitated release of
haemoglobin oxygen at high temperatures of the skeletal muscle tissue to allow more
oxygen release to the working muscles; (4) Higher temperature promotes faster
dissociation of haemoglobin and oxygen, a mechanism that helps to provide more oxygen
to active tissues; (5) Enhanced secondary nerve transmission, which could lead to
increased contraction speed and reduced reaction time at higher tissue temperatures; and
(6) increased blood flow through active tissue due to vasodilatation (Hoffman, 2002;
Woods et al., 2007). Increased skeletal muscle temperature and increased blood flow are
the main physiological changes due to warm-up, according to McHenry et al. (2009).
Most research has concentrated primarily on the physiologic characteristics of the
human skeletal muscles. There is a lack of studies that examine the characteristics of
laryngeal muscles in humans; therefore it remains unclear (Hoh, 2005). Skeletal muscle
fibre forms of laryngeal muscles have been observed in both human and animal models.
The presence of myosin heavy chain (MHC) isoforms characterizes skeletal muscle fibres
which regulates the speed of contraction (Barany, 1967). While various forms of skeletal
muscle fibre and hybrid combinations of different fibre types are likely to exist, the
classification will be summarized as follows for this discussion: Type I fibres are highly
Univers
ity of
Mala
ya
22
resistant to fatigue and contract slowly; Type II is the most fatigable but fastest-
contracting fibre types. Both types of muscle fibres vary concerning the bioenergetics
pathway used to form adenosine triphosphate (ATP). As an energy substrate, Type I
fibres rely on oxygen for the continuous supply of ATP. ATP is the usable energy form
of the body needed to initiate, sustain and stop muscle contraction. The fast-twitch fibres
can be further classified into three main subtypes: Type IIa, Type IIx, and Type IIb. Such
fibres mainly contribute to muscle speed and strength and differ in fatigability. For the
generation of ATP, immediate energy (phosphocreatine) systems and glycolysis at
different levels are provided by type IIx and IIb fibres (Hoh, 2005).
The laryngeal muscles of three healthy humans were examined by Teig, Dahl and
Thorkelsen (1978). The researchers stained the frozen portion of each muscle and
microscopically examined the fibres. The researchers concluded that the laryngeal
muscles consist of a combination of type I and type II muscle fibres. The research showed
that the Paired thyroarytenoid (TA) muscle has faster fibre type profiles than the Posterior
cricoarytenoid (PCA) muscle, which has faster fibres than the cricothyroid (CT) muscle
(Hoh, 2005). The PCA muscle is the main vocal cord abductor while the cricothyroid
(CT) muscle is the only tensor muscle of the larynx aiding with phonation (Deguchi et
al., 2011; Poletto, Verdun, Stominger & Ludlow, 2004). The variability in the distribution
of muscle fibre type in both skeletal limb muscle and laryngeal muscle differ from person
to person due to genetic predisposition. According to Hoh (2005), muscle endurance may
change to adapt to changes in function and hormonal stimuli, but the distribution of the
fibre type remains constant in a person.
A study by Brandon et al. (2003) found TA specimens to have types I, IIa, and
some had type IIx with more abundant type II (fast-twitch) fibres. Additionally, a few
rare fibres for both α-cardiac and type I were found. No muscle spindles were identified
Univers
ity of
Mala
ya
23
throughout the TA, which is in contradiction to a study by Han et al (1999). The
investigators suggest it is possible that, like extraocular muscles, human laryngeal
muscles use specialized spindle-like receptors and afferents (evolved specialized, non-
weight bearing movements). These receptors may provide feedback as to vocal fold
length and position as highly adapted kinesthetic receptors (Brandon et al., 2003b).
Kersing and Jennekens (2003) found almost equal proportions of muscle fibre
types I and II, and they acknowledged reports that the glycolytic and oxidative capacity
of the type II fibres (type IIa) indicate that the majority have the aerobic capacity, like
type I fibres. Therefore, both are “rich” in mitochondria. Like the Brandon study, no
muscle spindles were identified. A human TA research by Han et al. (1999) explored the
presence and distribution of slow tonic tone muscle fibres (STFs), the tonic of the
contractile properties of which differ significantly from slow as well as fast-twitch fibres.
Immunofluorescence microscopy revealed STF in the human TA.
The investigators found the STF to be localized in the middle of the vocalis
(anterior-posterior orientation). Additionally, these fibres were found interspersed
singularly or in small groups among twitch muscle fibres. The authors hypothesize that
such arrangements allow the STF to influence vocal fold vibration either directly or by
interaction with twitch fibres. These tonic fibres are thought to “exert a filtering effect on
the harmonics produced by vocal fold vibration (p.155)”. STF was also found interspersed
with muscle spindles. Together, these fibres may play a key role in proprioception (e.g.,
perception of position, length and tension). The authors conclude that “the presence of
many STF, or STF at certain locations within the TA, maybe one of the reasons that
certain individuals can sing professionally just as other genetic physical attributes allow
some people to be star athletes (p. 155)”. While the TA is comprised mostly of fibres of
oxidative metabolism, making it generally fatigue-resistant, it appears that there is
Univers
ity of
Mala
ya
24
considerable variation among individuals as to the proportion of types I and II and hybrid
muscle fibres. The current understanding of the effects of ageing on the TA as compared
to skeletal limb muscles was reviewed by Thomas, Harrison and Stemple (2008). The
authors conclude that the TA may be characterized as rapidly contracting and largely
fatigue resistant. This combination is unusual for limb skeletal muscle. Additionally, the
TA muscle differs from limb skeletal muscles in terms of its innervation, architecture,
mitochondrial content, contractile protein profile, and ageing patterns (Thomas et al.,
2008). The authors warn that the research concerning the typical skeletal muscle cannot
be easily generalized to the laryngeal musculature because of these variations. However,
generalized laryngeal fatigue remains debilitating for many averages, as well as,
professional voice users. Laryngeal fatigue is characterized by physical sensations similar
to that of skeletal muscles. Singers and speakers complain of muscle ache, soreness, and
tension, as well as, the need for increased effort to continue vocalizing. The physiological
and biomechanical phenomena associated with laryngeal fatigue remain under
investigation (Welham & Maclagan, 2003). As the skeletal muscles composition of the
vocal fold is being established, it is important to understand how the laryngeal muscles
can be affected by a vocal warm-up. Although it is difficult to directly quantify the impact
of vocal warm-up on the individual intrinsic laryngeal muscles, a well-established
standardized acoustic analysis tool can indirectly measure the collective influence of
warm-up (Milenkovic, 1987; Brown et al., 1990; Brown, Rothman, and Sapienza 2000).
The following section discussed the perturbation measures used in most research studies
to quantify the vocal quality.
Univers
ity of
Mala
ya
25
2.4 Perturbation Measures – Jitter, Shimmer & Harmonics-to-noise ratio (HNR)
Vocal perturbation measures quantify the average variability of the voice (Baken,
1990). Vocal perturbation measures are used typically to quantify vocal quality in the
speaking voice. Three vocal perturbation measures used by Milenkovic to quantify vocal
quality are jitter, shimmer, and signal-to-noise ratio (Milenkovic, 1987). Jitter is defined
as cycle-to-cycle variations in pitch. It is also known as the frequency perturbation or
modulation frequency. Shimmer, also known as amplitude perturbation or modulation
extent, refers to cycle-to-cycle variations in intensity. The Harmonics-to-noise ratio
captures the relative contributions of all periodic to aperiodic energy in the speech signal
(Teixeira, Oliveira, & Lopes, 2013; Bausar, Bohlender, Mehta, 2018).
Brown et al. (1990) compared jitter ratios of elderly female singers (63-85 years)
with young female non-singers (20-32 years) and elderly female non-singers (75-90
years). All subjects produced sustained phonation of the vowel /a/. No significant
differences were found. However, the elderly singers exhibited the lowest mean jitter ratio
values (0.22), compared with the young (0.27) and elderly non-singer groups (0.32).
Elderly singers and young non-singers had a standard deviation of 0.21 and elderly non-
singers 0.38.
The effects of isometric-isotonic vocal exercises practised daily for four weeks in
20 voice major graduate students (age range of 21-43 years) was studied by Sabol et al.
(1995). The experimental group and the control group was used in this study. The subjects
were randomly assigned to the two groups. The researchers decided to divide the groups
based upon the participants’ experience, level of singing, age, and sex (three males and
seven females in each group). The experimental group was taught how to properly
perform Dr. Joseph Stemple’s Vocal Function Exercises (VFEs), and the exercises were
added into their practice sessions, two times a day and twice per session. As previously
Univers
ity of
Mala
ya
26
completed, the control group continued their current practice sessions. All the participants
in both groups were required to continue their voice lessons and training and were told to
avoid verbal aggression (e.g., yelling) (Sabol et al., 1995).
Experimental and control subjects sustained the vowel /a/, /i/, /u/ at three pitch
levels. Pre-and post-test measures differed significantly at both high and low pitch levels.
For the experimental group, the mean jitter increased from 0.24 to 0.31 at the comfortable
pitch, while for the control group, the jitter decreased from 0.25 to 0.18. The jitter of the
experimental group increased at the low pitch from 0.26 to 0.32, while the control group
lowered by 0.01. These findings are surprising, and in contradiction with the previous
studies. One would expect that with vocal exercise the irregularities found in vocal fold
function due to variation of mass, tension, muscle or neural activity would be reduced.
Sabol et al (1995) evaluated the effects of Dr. Joseph Stemple’s Vocal Function
Exercises (VFEs) on vocal measures when they were integrated into singers’ everyday
practice. It was a pre-test and post-test group control design. Acoustic measures (jitter,
frequency range and fundamental rate), aerodynamic measures (phonation
volume, maximum phonation time and flow rate) and video stroboscopic measures were
evaluated.
Sabol et al. (1995) obtained acoustic and aerodynamic analysis data using a Visi-
Pitch (Kay Electrometrics model 6097) and a Nagashima Phonatory Function Analyzer
(PS 77H model). Additionally, data were collected through questionnaires and case
histories that revealed the participants’ extracurricular activities, voice type, medical
history, and schedule of singing. Similar to Stemple et al. (1994), the researchers
conducted the study for 4 weeks, collecting data 28 days apart for pretest and posttest data
collection.
Univers
ity of
Mala
ya
27
Sabol et al. (1995) completed a statistical analysis of the data by utilizing a
multivariate analysis of variance (MANOVA) and an ANOVA to assess for significance
through a p-value. Significant variations between pre- and post-test findings in phonation
volumes, lower airflow rates, and higher phonation times was found in this study. This
study revealed higher phonation volumes to that of Stemple et al. (1994) with non-trained
singers and required further investigation as to an explanation. In regards to the decreased
airflow rates, it was suspected that VFEs assisted in the enhancement of vocal fold
adduction, adequate phonation, and subglottic pressure. For maximum phonation time, it
was thought to have increased due to an easy onset of phonation with proper glottal
closure at low lung volumes resulting in an increase of phonation time (Sabol et al., 1995).
Sabol et al. (1995) demonstrated that vocal function exercises (VFEs) had a
beneficial effect on trained singers without a voice disorder. These authors concluded that
further research is needed to evaluate VFEs on younger singers who are in their
undergraduate program and who are not as advanced as graduate trained singers are. This
study showed, however, that even advanced trained singers benefited from the use of
VFEs and that they could use them as a long term, practice regimen.
Later on, Brown, Rothman, and Sapienza (2000) compared jitter percentages
produced by 20 professional singers and 20 non-singers during spoken and sung vowels.
Subjects sustained the spoken and sung /i/ vowels of the word “sea” of “America the
Beautiful”. The group consisted of 10 men (22 to 62 years) and 10 women (21 to 61 years
of age). The age-matched group of non-singers consisted of 10 males and 10 females,
both ranging from 22-63 years. Jitter percentages for the spoken /i/ were significantly
higher for the male non-singers than the male singers. The female singers had a higher
mean jitter than the non-singers did, but no significant differences were found. Overall,
jitter percentages for the sung /i/ were higher for non-singers than singers.
Univers
ity of
Mala
ya
28
Mendes et al. (2004) found that singing warm-ups and training had an important
impact on singing, especially in raising both the fundamental frequency and intensity of
the phonation range. They studied 14 professional singers over four college semesters to
measure the effects of singing training. Subjects were recorded after briefly warming-up
their voices by performing dynamic range exercises to expand and stretch their vocal
range. They then sustained vowels and read the “Rainbow Passage” while being recorded.
At the end of the study, jitter and shimmer measures slightly decreased, and phonation
range increased. The measure of phonation range is also referred to as maximum
phonation frequency range (MPFR), and it is defined as the range of frequencies of both
the modal and falsetto registers from the lowest sustainable tone to the highest (Mendes
et al., 2003). In studies that have compared singers to non-singers, singers typically have
a larger MPFR and are, therefore, able to sing a larger range of tones and can stretch their
voice between multiple octaves (Mendes et al., 2003). Thus, vocalization exercises can
extend the vocal range the same way as stretching before a physical exercise (McHenry
et al., 2009).
Shimmer, or cycle-to-cycle amplitude variation, is another perturbation measure.
As was the case for the measurement of frequency perturbation, the amplitude of each
glottal cycle is measured, subtracted from the previous or following period, and averaged
over all differences. This is called the shimmer ratio (Horii, 1980; Lathadevi & Suresh,
2008).
Brown et al. (2000) studied shimmer measurements from speaking and singing
tasks between male and female singers and non-singer groups. Each group had 10 subjects
with ages ranging between 21 to 63 years. The singer group are comprised of ten males
(22-62 years) and ten females (21-61 years). The age-matched non-singer group
comprised of ten males and ten females both ranging in age from 22-63 years. Subjects
Univers
ity of
Mala
ya
29
sustained the spoken and the sung /i/ vowels of the word “sea” from the song “America
the Beautiful” for five seconds each. Shimmer values for the/i/ spoken showed no major
variations between singers and non-singers. For the sung /i/, the shimmer values of non-
singers were significantly higher than those of the singers. Harmonics-to-noise ratio
(HNR) is the noise-related measure which is just like a ratio or expressed in dB. The voice
is a quasiperiodic instrument with two components: quasiperiodic waves and random
noise. The harmonics-to-noise ratio (HNR) is the mean amplitude of the average divided
by the mean amplitude of the isolated wave noise components (Baken, 1987; Narasimhan
& Nataraja, 2019).
Comparative studies of HNR with the singing population, as for shimmer, are rare.
Brown et al. (2000) compared the same singer and non-singer subjects during the
sustained spoken and sung /i/ vowels for HNR. The authors reported no significant
differences between singers and non-singers for the speaking sample. For the sung /i/
sample, male singers showed a significant larger HNR than the non-singer group. No
differences were found in the female group.
Similar research was carried out by Tay et al. (2012) to investigate the impact of
VFEs on vocal function measurements. It was a prospective experimental design which
comprised of 22 ageing choral singers. The vocal measures evaluated were acoustic
parameters (jitter, shimmer, phonation frequency range, and noise-to-harmonic ratio
(NHR)), maximum phonation time and auditory-perceptual voice features such as
breathiness, roughness, and strain. There were two groups, which are the experimental
vocal function exercises (VFEs) or the control group, in this study and the subjects were
assigned into both groups randomly, each of which contained eleven participants. To
ensure that vocal function exercises (VFEs) was correctly conducted during the study, the
control group had to meet the researcher twice. During the seven weeks, the experimental
Univers
ity of
Mala
ya
30
group participants had to attend four sessions with the researcher. Both experimental and
control groups had to record their singing hours and practice times.
Tay et al. (2012) used the Mann Whitney U test and a paired t-test to complete
statistical data analysis. No statistical difference was found between the two groups.
Nevertheless, the results of the Mann-Whitney-U test revealed that the control group sang
more hours than the experimental group during the study. With regards to the breathiness,
roughness and strain, only roughness showed a statistically significant decrease for the
experimental group. It is suspected that this was due to the improvement of the vibration
of the vocal folds since performing vocal function exercises (VFEs). The results showed
increased phonation frequency range in the experimental group and decreased jitter,
shimmer, and HNR in both groups. It was probably because more singing hours were
recorded by the control group than by the experimental group. Furthermore, it was
believed that an increased frequency range of phonation and a decrease in jitter and
shimmer showed that these exercises improved vocal fold adduction and increased
vibration. Tay et al. (2012) gave a self-evaluation consisting of a scale of 0 to 6, 0
represented strongly disagreed and 6 represented strongly agree. The subjects were told
to determine if their voice had improved after the study if their voice had changed as
VFEs were utilized and whether they would continue the vocal function exercises (VFEs)
on their own. Results of the self-assessment showed that the participants in the
experimental group reported that their voices fatigued less and that they could sing longer
after performing VFEs. Overall, the study concluded that the experimental group's use of
VFEs improved overall voice quality.
Univers
ity of
Mala
ya
31
2.5 Acoustic Voice Analysis Using Praat Software
Praat is a powerful acoustic analysis software written by Paul Boersma and David
Weenik of the University of Amsterdam (Paul & Vincent, 2001). It is a freeware available
for download via the Internet, and has research literature validating its use for acoustic
analyses (Lathadevi & Suresh, 2008; Bausar, Bohlender & Mehta, 2018).
Lathadevi & Suresh (2008) used Praat software to analyze objective acoustic
analysis determine whether objective acoustic analysis is useful in differentiating
abnormal and normal voices. They found that abnormal voice parameters such as jitter
(ddp), shimmer (dda), average pitch and measurement of HNR were different from
normal voices. Significant differences were found in jitter males and shimmer females
but HNR has shown no significance. They concluded that the acoustic parameters allow
clinicians to characterize their voice into either normal or abnormal voices. Nonetheless,
a standard local or regional normal voice database is important for comparison.
Bausar, Bohlender & Mehta (2018) examined the impact of voice sound pressure
level (SPL) on acoustic parameters in 58 female voice-disordered adults and a control
group of 58 vocal health women. Praat software was used to compute acoustic voice SPL,
jitter, shimmer, and HNR. The increased voice SPL was significantly associated with
decreased jitter and shimmer (P < 0.001) and increased HNR in both patient and
normative control groups. Professional voice use level had a significant effect (P < 0.05)
on jitter, shimmer, and HNR. They concluded that as voice intensity increased, acoustic
perturbation improved, with jitter and shimmer decreasing, and HNR increasing.
Univers
ity of
Mala
ya
32
2.6 Physiological Effects of Vocal Warm-up
Phonation threshold pressure (PTP) is an indirect measure of vocal fold function.
It can be used to determine how vocal warm-up can affect the physiological
characteristics of the vocal folds. According to Titze (1988 & 1992), PTP represents the
minimum subglottal pressure required for vocal fold oscillation. PTP is a measure of ease
of perception and phonation efficacy that is directly related to the viscosity of the vocal
fold mucosa (Titze, 1988; Chan & Titze, 1998). When vocal warm-up enhances the vocal
fold conditions physiologically, PTP is decreased (Milbrath & Solomon, 2003). A handful
of studies have observed the relationship between a warm-up condition and phonation
threshold pressure as a variable (e.g. Motel, Fisher, & Leydon, 2003; Elliot, Sundberg, &
Gramming, 1995). Theoretically, the vocal warm-up will increase blood flow in muscles,
decrease viscosity in the vocal folds and thus lead to a lower phonation threshold pressure
(Milbrath & Solomon, 2003).
McHenry et al. examined the impact of a typical vocal warm-up versus one
coupled with an aerobic routine to determine if changes in laryngeal muscle viscosity
could be increased when coupled with aerobics. They examined 20 participants, both
male and female, and discovered that aerobic routine did not necessarily affect the vocal
folds directly. They were all familiar with the vocal warm-up. However, both warm-up
activities created a general perception of less effort. Females also show that phonation
threshold pressure decreases significantly following a warm-up in this study.
Another benefit to a vocal warm-up is a decrease in overall tension. According to
Elliot, Sundberg, and Gramming (1995), the temperature of the muscle is increased after
a warm-up, and thus muscle viscosity is reduced. It is reasonable to assume that vocal
warm-up will have the same effect on the laryngeal muscles. Koufman, Radomski,
Joharji, Russell, and Pillsbury (1996) studied the patterns of laryngeal tension on the
Univers
ity of
Mala
ya
33
singing voice in 100 healthy singers via transnasal fiberoptic laryngoscopy and found that
both trained female professional singers and classical singers had very low muscle tension
scores. The significance of this finding is that excess muscle tension may be the cause of
phonotrauma and other voice disorders so that an increased training in vocal therapy and
warm-up activities before singing can only benefit individuals by reducing laryngeal
tension (Stemple et al., 2010)
The phonation threshold pressure (PTP) effects of soprano singers in a 10-minute
vocal warm-up were measured by Motel et al. (2003). In this study, PTP was obtained at
10 per cent, 20 per cent, and 80 per cent of the total frequency ranges of ten pre- and post-
conditions female voice majors: vocal warm-up and vocal rest. It lasted for 10 minutes
and the exercise was selected to resemble the traditional warm-up procedure at the start
of a daily voice lesson. Audio recording, including the accompaniment of the piano,
guided the participants. Findings showed that at the 80 per cent frequency, vocal warm-
up increased PTP. Higher PTP is associated with increased effort. PTP increased
following a vocal warm-up is a result of muscular force production (Titze 1988). As it
was reported that the TA muscle is fast-fatigable (Teig et al., 1978), the chances were
high that 10 minutes of vocal warm-up led to increasing in PTP due to TA muscle
fatigued. Nevertheless, the participants’ perception results showed that some participants
were pleased with the warm-up and one of them indicated that they felt ready for
performance (Motel et al., 2003). In another research investigated by Elliot et al. (1995),
phonation threshold pressure was obtained from ten amateur singers, before and after a
warm-up. The warm-up of approximately 30 minutes. The singers sang/mu/ on a
descending scale followed by other pitch-changing exercises with various vowels and
different levels of vocal loudness. All the participants were informed not to sing
extremely loud during the warm-up. All participants reported verbally that they felt they
Univers
ity of
Mala
ya
34
had better voices after the warm-up. They indicated that singing was easier, especially at
high pitches, and they felt like having more control over their voices while singing. This
research was conducted to compare the participant's subglottal pressure the fundamental
frequency curve with Titze curve (1992) indicates the decreased thickness of the vocal
fold correlated with an increase in fundamental frequency. The male participant curves
were fairly close to the theoretical curves. Phonation threshold pressure (PTP) seemed
consistently pitch-dependent for male participants, but there was no similar pattern in
female participants. No consistent results have been found. There was no post-warm-
up trend in PTP, with some of the participants showed increasing PTP, some decreasing
PTP, and some not significantly changing PTP.
Milbrath et al. (2003) studied the effect of vocal warm-ups on phonation threshold
pressure (PTP) on eight female non-singers who have voice fatigue complaints. The
researchers investigated the vocal warm-up effect on perceived phonatory effort (PPE).
PTP and PPE data were collected before and after two “vocal-preparation” conditions:
(1) 20-minute vocal rest and general relaxation, and (2) 15-20 vocal warm-up routines
created by researchers. This warm-up exercise involved stretching legs, neck, back,
mandible and tongue, and then deep breathing while seated with a well-aligned posture.
After that, the focus was on resonance as participants held pitches for one to two seconds
while concentrating on “feeling the upper lip or nose buzz”. The participants were finally
instructed to perform vocal exercises consisting of upward and downward pitch glides.
PTP and PPE were measured after a vocal loading task of one hour (loud reading) and
again after a vocal recovery time of 30 minutes following the vocal preparation
conditions. For each of these conditions, there were no significant differences between
PTP and PPE. The researchers found that PTP and PPE are not sensitive to determining
vocal function changes in people with chronic vocal fatigue.
Univers
ity of
Mala
ya
35
As measured by phonation threshold pressure (PTP) and perceived phonatory
effort (PPE), performing warm-up as a method to avoid fatigue or minimize effort
resulted in inconsistent results. Empirical findings suggest that after vocal loading, vocal
warm-up exercises do not seem to have a positive impact on PTP or PPE and there is a
possibility that PTP rising following a 10-30 minute vocal warm-up (Motel et al., 2003;
Elliot et al., 1995). Given the conceptual effects of increasing PTP, increased effort in
strictly controlled intensity and frequency, PTP increased after vocal warm-up (Motel et
al., 2003; Elliot et al., 1995). According to Motel et al. (2003), phonation ease following
the vocal warm-up was reported as the phonation threshold pressure increases. There are
two plausible reasons. One explanation is that increased PTP at high pitches is correlated
with positive phonatory changes due to an ischemic effect that prevents high-frequency
vascular damage or loss of mucosal fluid. Another reason is that increased water
efflux from vocal warm-up regimens in the vocal fold may help to facilitate phonation.
Milbrath and Solomon (2003) also found little evidence for warm-up benefit through PTP.
The researchers carried out experiments on warm-up exercises and loud reading but
results indicated that the experiments were not sensitive enough to vocal-function
changes. Elliot et al. (1995) propose that phonation threshold pressure variation may be
due to complex properties of the vocal folds or individual differences following the vocal
warm-up.
Univers
ity of
Mala
ya
36
2.7 Vocal Tract Tuning, Resonance and Acoustic
Vocal warm-up is also important for vocal tract tuning. It can affect the changes
in supraglottal and glottal settings and therefore changes the vocal quality. Optimal
resonance can be achieved through vocal warm-up. Laukkannen, Horacek, and Havlik
(2012) conducted a study whereby they used magnetic resonance imaging (MRI) to
examine a male and a female to determine post-warm-up vocal tract changes. The
participants were recorded twice, before and after a “some minutes” personalized vocal
warm-up. Each participant was instructed to sustain each vowel, /a/, /i/ and /u/ for 20
seconds. To describe vocal tract size and shape, Midsagittal MRI images were analyzed.
In particular, the pharyngeal inlet over the epilaryngeal outlet ratio was determined (Aph
/ Ae). The main change in males was that the larynx decreased from fifth to sixth vertebrae
after warm-up. No vertical shift was found in the female larynx following the vocal warm-
up. There was only a marginal rise in vowel /u/. A certain opening of the jaw was observed
when the participants sang /a/. The tongue was frontal and curved, and the pharynx
became wider after the warm-up. Aph / Ae increased in both groups by up to 27 per cent in
males and up to 28 per cent in females for all the three vowels. Despite the small sample
size, this research indicates that supraglottal changes can be observed after the vocal
warm-up. The sound produced by vocal folds, which can be interpreted as a filter
characteristic, is generated by supraglottal structures (Fant, 1970). Resonance occurs
during vowel production when sound waves passed through the vocal tract. The vocal
tract is a natural cavity resonator. Resonance is defined as the reinforced, natural
oscillation of the object (Titze, 2000). The harmonic spectrum frequencies are modified
by the resonance of the vocal tract. In the supraglottal structures, specific frequencies
from the harmonic spectrum resonate, while others are damped by soft tissue. Formants
are frequency peaks in the spectrum which have a high degree of energy (Fant, 1970).
Univers
ity of
Mala
ya
37
The spectral envelope is a smooth curve that surrounds the amplitude spectrum (Fant,
1970). The specific frequencies of formants are placed on the spectral envelope using
linear prediction code (LPC) assessment (Hixon et al., 2008). F-patterns, named F1 (first
formant), F2 (second formant), and F3 (third formant) are the first three peaks in the
spectral envelope (Fant, 1970). The F-patterns can infer the articulation or filter shape.
The F1 (first formant) typically increases with decreased tongue height, the second
formant, F2 increases with increased advancement of the tongue and as the opening of
the mouth increases in size and become less rounded, the third formant F3 increases
(Naderifar et al., 2017). The vocal warm-up, as well as the modifications to the vocal
tract, can determine the resulting acoustic qualities (Laukkanen, Horacek, & Havlik,
2012). Laukkanen, Horacek, and Havlik (2012) considered the ratio of pharynx inlet and
epilarynx outlet region is important in the formant production, which improves the
acoustic voice qualities. The singer perceives the sound during singing and warm-up
through both the vibratory sensations in the head, neck and chest and the auditory system,
which are felt when the singer manipulates their instrument to achieve maximum
resonance (Vurma & Ross, 2003). This practice of resonant voice has been found to have
a production that feels easy and adequately loud, while also being felt vibrating within
the facial tissues (Ziegler, Gillespie & Verdolini Abbott, 2010). The singer can physically
feel the vibration of resonance as they warm-up and sing. The vocal tract has four or five
major resonances or formants. The shape of the vocal tract determines the formant
frequencies. Trained singers learn to manipulate these formants by manipulating the
hypopharynx and the epilarynx (Sundberg, 1977). By lowering the larynx, the lowest part
of the pharynx is expanded, as is the laryngeal ventricle, the space between the true and
false vocal folds (Sundberg, 1977). Increasing the width and shape of the vocal tract leads
to changes in resonator or vocal tract structure, and these variations in resonance are the
Univers
ity of
Mala
ya
38
most noticeable differences in the singing voice as compared to the speaking voice. The
increases in width and length of the vocal tract result in an increase of resonant energy
between 2500Hz and 3000Hz, which is also known as the “singer’s formant” (McCrea &
Morris, 2007). The singer’s formant corresponds to the “ringing” quality of the singer’s
voice and this is often referred to as “getting resonance,” as a resonant voice creates
changes in the vocal tract, which create a maximum transfer of power through the vocal
tract and out to the listener (Titze, 2004). Resonance can also be increased by widening
the pharynx during singing while also lowering the larynx, a sensation that is similar to
yawning. As the throat is “opened” by these actions, this yawning sensation can aid in
stretching the vocal muscles, readying them for optimum resonance and phonation
(Sundberg, 1977). Warm-up exercises are generated by resonance effects and enhancing
the resonance of the vocal tract, which therefore reduces any need for extra laryngeal
effort, helping the singer to achieve maximum sound and intensity without having to
generate extra pressure (Sundberg, 1977).
Warm-up exercises often focus on the idea of the forward focus resonance to
ensure that vowels are “placed” correctly and to reduce the excessive muscle tension that
can be found with a “backward” voice, which often diminishes the effective use of the
resonators (Vurma & Ross, 2003). In accomplishing forward focus, optimal resonance
can be achieved which may also enhance the singer’s formant or the ringing quality of
the singing voice.
Many types of vocal therapy exercise also touch upon this type of “warm-up,”
which focuses on the benefit to resonance and proper projection of resonance for optimal
voice output. The vocal function exercises of Stemple aim to balance the voice production
subsystem and demand coordination of the respiration, laryngeal movement, and
resonance so that a maximal voice output is achieved with minimal effort (Stemple et al.,
Univers
ity of
Mala
ya
39
2010). Opera singers, natural voice speakers and voice condition teachers have been
studied and it has proved effective to improve and enhance their voice function (Stemple
et al. 2010). Resonant voice therapy is similar in its conceptualization. This therapeutic
warm-up “forward focus” which was derived from performing arts training, is described
as voice production involving the vibratory sensations in the oral cavity and face in the
context of easy phonation. By enhancing these vibratory sensations while also practising
easy phonation, the goal becomes achieving the strongest possible voice with minimal
effort and vocal fold impact. Resonant voice tends to be associated with vocal folds that
barely touch, which optimizes the relationship between a large voice output volume and
very little vocal fold impact (Ziegler, Gillespie & Verdolini Abbott, 2010).
Miller, Sulter, Schutte, and Wolf (1997) found through magnetic resonance
imaging (MRI) of a singer’s vocal mechanism during various types of singing, that certain
postures of the vocal tract were adjusted, which enhanced elements of the singing voice,
such as resonance, and the singer’s formant. Trained singers often sing better than non-
singers because those trained singers have learned a series of phonatory, articulatory, and
resonatory adjustments that non-singers do not use. The ability to have a skill in some
sort of performance is often attributed to talent. The natural ability to produce musical
changes in the voice, where frequencies are in tune with one another and sound melodic
is defined as singing talent (McCrea & Watts, 2007).
Univers
ity of
Mala
ya
40
2.8 The Singer’s Formant
This section discusses the importance of vocal warm-up to increase the intensity
of the singer’s formant. The singer’s formant indicates prominent sound energy near 3
kHz, mainly found in the voice of classical and operatic singers (Sundberg, 1974;
Sundberg, 1998). In female voices, higher frequency spectral reinforcement is much less
established, therefore the singer's formant is usually found in male voices (Morris &
Weiss, 1997). Studies of classically trained singers and especially males showed a clear
and strong formant of approximately 3000 Hz (between 2800 and 3400 Hz) which is
absent in the voices of untrained singers. The singer’s formant is more prominent in solo
singing than choral singing (Sundberg, 1998). This formant reflects the frequency
spectrum of trained classical singers and the resonant frequency has been designated as
wide-spread as 2600 to 4000 Hz (Morris &Weiss, 1997). The presence of the singer’s
formant is to support the voice and make singers heard and understood through a
full orchestra (Sundberg, 1998).
A formant cluster happens when the formant amplitudes increase and the third,
fourth and fifth formants approach each other in a frequency (Sundberg, 1974; Sundberg,
1998), a formant is generated. A cluster of formants was not found in female voices. Only
strong upper-frequency reinforcement was observed in female voices (Morris & Weiss,
1997). Sundberg suggested that since the partials of higher frequencies are well above the
strongest orchestral sounds, the risk of masking by an orchestra is slight, therefore
singer’s formant cluster in female voices is not necessary (Sundberg, 1974). The
articulatory formation of the singer’s formant comprises specific parts of the vocal tract
(Sundberg 1974). The way to modify vowels such as by movement of the tongue, lips and
adjustments in laryngeal height, either enhances or inhibits the sound. The tweaking of
vowel colour associated with adjustments in F1 and F2 (referred to as “formant tuning”
Univers
ity of
Mala
ya
41
or “vowel tuning”) is a big part of how singers can learn to optimize the resonance. When
the larynx is lowered, increased formant density usually occurs (Sundberg, 1974; Morris
& Weiss, 1997; Sundberg, 1998). Sundberg (1974) claimed that the lowering of larynx
leads to the singer’s formant. There is the widening of the laryngeal ventricle and
pyriform sinuses. Next, by lowering the larynx, the vocal tract’s length is raised. It is
essential to lower the larynx to make separate but complementary adjustments of the
pharynx and epilarynx tube pharynx to form the singer’s formant.
The epilarynx tube is small, about 2 cm in length, situated above the vocal folds
(Sundberg, 1974). The larynx is located directly above the pharynx, a larger tube above
the epilarynx tube that forms the rest of the throat (Sundberg, 1974). The narrowing of
the epilarynx tube and reducing the epilarynx tube opening to less than 1:6 was thought
to be connected to the articulatory basis for the singer's formant (Sundberg, 1974). The
pharynx expands as the epilarynx tube narrows to help achieve this ratio (Sundberg, 1974;
Titze, 2012). The exercises that facilitate the widening of the pharynx (Titze, 2012) and
lowering of the larynx (Weiss, Brown & Morris, 2001) are important training to develop
the singer's formant (Titze, 2012). The singer's formant is typically found in classically
trained male singers (Sundberg, 1998; Morris & Weiss, 1997; Sundberg, 2001). Morris
& Weiss (1997) claimed that without proper training, the singer’s formant can hardly
emerge, although most vocal coaches consider this to be the by-product of good singing.
Besides, if the frequencies are very high, the singer’s formant can hardly be produced, for
instance over a C5, the presence of the formant is infrequent. High-frequency harmonics
are further separated and peaks often hard to detect, especially where a partial near-
formant frequent is not available (Sundberg, 1998). Therefore, in male voices, it is more
accurate and more discernible to observe and measure the singer's formant than in female
voices.
Univers
ity of
Mala
ya
42
Warm-up exercises have been found to increase the intensity of the singer’s
formant, resulting in a greater ‘ring’ to the singing voice. A “ringing” quality in the voice
is the perceptual result of these vocal tract adjustments (Omori, Kacker, Carroll, Riley, &
Blaugrund, 1996; Elkholm, Papagiannis, & Chagnon, 1998). Vinturri et al. (2001) studied
the influence of vocal warm-up on the singing voice. Individuals were selected randomly
in the study. The subjects were all non-singers who had not gone through formal voice
training or voice therapy. They found that after a warmup, the spectral energy in the
singer’s formant region between 2500 Hz and 3000 Hz increased significantly in all tested
pitches among the female subjects. Their participants also noted that after a warm-up they
felt their voice was easier to use and felt smoother in both singing and speech. In this
study, the impact of vocal warm-up on non-singers was explored, their warm-up strategies
were focused on intensity changes, rather than a more general vocal warm-up, and thus
these results are limited in their ability to predict the effect on non-singers.
Univers
ity of
Mala
ya
43
2.9 Measurement of Singer’s Formant
According to Sundberg (1995), when the intensity of the formant is “exceptionally
high” near 3 kHz, the absence or existence of the singer's formant can be established. The
method used to calculate the exceptionally high-intensity level (the level for Sundberg) is
to compare the level of the third formant (L3) to the predicted level of L3 based on the
equations of Fant (1970), to estimate the level of the formant. The existence of the formant
of the singer is indicated by the difference between the observed and the predicted L3
being greater or equal to 6 dB. Vowels influence the relative L3 when this method is used,
and therefore the singer's formant level varies with the vowel (Sundberg 1995).
Another method of measuring harmonic peaks at the upper frequency is the
singing power ratio (SPR). SPR is used to objectively differentiate trained singers from
untrained singers (Omori et al., 1996; Amir, Amir, & Michaeli, 2005; Watts, Barnes-
Burroughs, Estis, & Blanton, 2006). The ratio of the highest intensity peaks between 2-4
kHz and 0-2kHz range, is expressed in dB, is known as the Singing Power Ratio (SPR)
(Omori et al., 1996). Singing power ratio provides an implicit finding that vocal warm-
ups can affect the harmonic tuning in the vocal tract (Laukkanen, Horacek & Havlik,
2012; Watts et al., 2006).
Amir, Amir, and Michaeli (2005) studied 20 female classical singers’ pre-and-
post- vocal warm-up. While the singers sustained vowels at 20%, 50%, and 80% of their
frequency range, the authors performed acoustic measures on their vocal samples. They
found that all frequency and amplitude perturbation measures, noise-to-harmonics ratio,
and singer’s formant measures improved significantly after warm-up. Jitter, relative
average perturbation (RAP), and shimmer all decreased significantly after warm-up.
Noise-to-harmonics ratio also decreased significantly, although still within normal limits,
from 0.101 to 0.096 after warm-up. They also found that singing power ratio (SPR)
Univers
ity of
Mala
ya
44
decreased significantly, from -29.25 to -27.82 following warm-up. Singing power ratio
(SPR) correlates directly to the singer’s formant intensity. The resonant tuning is
measured in the vocal tract and the SPR shows its acoustic characteristics (Omori, Kacker,
Carroll, Riley, and Blaugrund, 1996). SPR is determined by measuring the peak strength
ratio between 2-4 kHz and 0-2kHz as sustained vowels or vocal segments. The lower the
measurement of the SPR, the more energy that has been shown in higher harmonics
corresponds to the perception of a “ring” in the singing voice. Omori et al. (1996) found
that the SPR of a singer's sample was significantly lower compared to non-singer,
suggesting that singing efficiency can be quantified by calculating SPR.
Univers
ity of
Mala
ya
45
2.10 Disordered Voices
Blaylock (1999) stated vocal warm-ups are essential to maintain a healthy voice
of speech and singing, and users could benefit from structured vocal warm-up routine
based on physiology of voices. A variety of problems can arise for both speakers and
singers if the vocal mechanism is functioning inadequately. Singers can be greatly
affected if they sing professionally as a career. Singers with all levels of training are often
more sensitive to tiny changes in their voice (Cohen, Noordzij, Garrett & Ossoff, 2008).
Cohen et al. (2008) created a singer’s voice handicap index that modified elements
of the Voice Handicap Index (Jacobson, Johnson, Grywalski, Silbergleit, Jacobson,
Benninger & Newman, 1997) to create a questionnaire that focused more on areas that
may affect a singer. Singers are often at a higher risk for vocal problems and also perceive
impairment much sooner and differently than a non-singer (Cohen et al., 2008). Singers
will notice small changes in their voice and will often seek medical care more quickly,
especially if they sing professionally. They often have lower Voice Handicap Index
scores, showing less voice handicap than non-singers (Cohen et al., 2008). For singers
having vocal difficulty in the form of nodules or other voice disorders, the positive
benefits of vocal therapy for various voice disorders has been broadly researched.
However, the use of singing is not typically involved in these techniques. Thus, the need
for a positive therapy technique that is more dedicated to the small changes in the singing
voice, such as a vocal warm-up, could benefit these individuals (Cohen et al., 2008).
Beyond its function for singers, the clinical relevance of vocal warm-up is
substantial as it can potentially benefit not only professional singers and non-professional
singers who wish to improve their voice quality, but also the non-singing voice disordered
population. The vocal warm-up is a beneficial means of therapy in individuals with voice
disorders (Chernobelsky, 2007). Blaylock (1999) closely studied 4 subjects with voice
Univers
ity of
Mala
ya
46
disorders, all of whom had a vocal pathology, such as nodules or hoarseness that had
shown no major improvement with voice therapy or medical intervention. He taught them
a systematized vocal warm-up after they had stopped voice therapy and asked them that
they practice it once per day. The warm-up consisted of a sequence of vocal exercises that
adjusted tempo, vowels, range, volume, and intensity that worked to increase
coordination, flexibility, and increased strength of the vocal musculature. He assessed the
subjects every three weeks after a warm-up and asked them to sing, sustain vowels, and
speak. His results showed improvement. The more the participants performed a vocal
warm-up, the better their voice improved over time. Blaylock (1999) examined the levels
of increased intensity and waveform consistency, as well as the enhancement in harmonic
information through the analysis of waveforms, sonograms and spectrograms, all of
which showed improvements in the subject's voice samples over time. Singing teachers
were also instructed to rank the samples based on their subjective experiences to reflect
what they considered an increase in overall voice quality and these rankings also showed
improvement in perceived voice quality. Lastly, subjects also provided subjective reports
of their feelings about their voice quality. They reported their voice quality improved
following the vocal warm-up, and good voice quality lasted longer as they continued their
vocal warm-up exercises.
Voice therapy with professional singers and non-singers with voice disorders has
also been shown to be an effective tool to reduce vocal nodules, which often affect singers
(Chernobelsky, 2007). Therefore, having seen the positive benefits of a singing warm-up,
individuals may benefit from its use in all areas of voice use, from the disordered
population to improve deficits, to singers enhancing their craft, and to non-singers who
have a passion to sing and also want to improve their singing abilities.
Univers
ity of
Mala
ya
47
2.11 Vocal warm-up
According to Motel et al. (2003), the vocal warm-up is considered as necessary
for optimal voice. Warm-ups are used to address specific vocal issues and
repertoire problems, and to address the fundamental elements of good vocal technique
(Hylton, 1995). The singers reported more comfort when singing following the vocal
warm-up, but little is found regarding the impact of warm-up in voice production (Amir
et al., 2005; McHenry et al., 2009; Elliot et al., 1995; DeFatta et al., 2012, & Vinturri et
al., 2001).
Vocal warm-ups have also been shown to help reduce the thickness of the vocal
folds, change the velocity of the surface waves and modify the width of the glottis before
phonation. These changes affect many characteristics of the vocal mechanism, which then
affect voice production (Amir, Amir, & Michaeli, 2005). In principle, vocal warm-up
could improve vocal tract performances (Laukkanen, Horacek, & Havlik, 2012), prevent
vocal fold muscle fatigue (Milbrath & Solomon, 2003; Motel et al., 2003), physically
warm-up the vocal folds (McHenry et al., 2009), or improve vocal efficiency (Laukkanen,
Horacek, Krupa & Svec, 2012). There was no evidence of vocal warm-up causing
physiologic changes to vocal folds (Milbrath & Solomon, 2003; Cooper & Titze 1985;
Motel et al., 2003; Elliot et al., 1995) but vocal tract and subsequent vocal acoustic
changes were found in research studies (Laukkanen, Horacek, Krupa & Svec, 2012).
Physical warm-up exercises can range from passive warm-ups, such as massage, to
general warm-ups, such as jogging, to specific warm-ups involving actual activity
movements (McArdle et al., 1996). The content and duration of vocal warm-up routines
among singers and professionals varied, too (Sims and McWhorter, 2012; Gish, Kunduk,
Milbrath and Solomon, 2003), though most are “variations of a few key topics” (Titze,
2001). Singing teachers are facing challenges like how to warm up (Barr, 2009). The
Univers
ity of
Mala
ya
48
warm-up time varies considerably in particular between the vocalists (Gish et al., 2012).
For instance, a 20-minutes vocal warm-up was suggested by Miller, (1990). Nonetheless,
in some studies, warm-up protocols lasted from 7 to 30 minutes (Gish et al., 2012). Most
vocal warm-up exercises lasted from 5 to 10 minutes reported by a study by Gish et al.
(2012). According to Elliot et al. (1995), an ideal vocal warm-up time duration ranging
from 10 to 30 minutes enhances voice performance for most singers. Miller (1990),
however, warns that singing for more than 30 minutes will affect the voice production
quality. In designing a proper definition of vocal warm-up, both of these opinions were
taken into consideration: traditional vocal warm-up and semi-occluded vocal tract warm-
up.
Univers
ity of
Mala
ya
49
2.11.1 Traditional Vocal Warm-up & Semi-Occluded Vocal Tract Warm-up
The research detailing vocal warm-up routines provides various exercises and
sometimes conflicting evidence (Gish et al., 2012; Milbrath & Solomon, 2003 & Barr,
2009). Vocal warm-up began with exercises designed to stretch, loosen and helped ease
the specific muscles involved in singing was proposed by Hylton (1995). Before vocal
warm-up, a brief aerobic exercise was also suggested by (Miller, 1990) with the idea of
the voice is a full-body instrument (Shear, 2008). The differential effect of “combined”
warm-up and specified vocal warm-up was investigated by McHenry et al (2009).
Researchers have found that aerobic exercises had a stronger impact than a specific vocal
warm-up on the core body's temperature. The vocal fold composition, the
unanticipated differences in physical activity levels, or hyaluronic acid difference in the
vocal folds were attributed hypothetically to differences in results between men and
women. Evidence indicated that, among women of average fitness, physical activity
coupled with a 20-minute vocal warm-up resulted in a greater decrease in post-warm-up
PTP (McHenry et al., 2009). The decline in PTP measured shows a lower viscosity and
thickness of the female vocal folds, which made the initiation of the phonation easier
(Titze, 1988).
Following light physical exercise or diaphragmatic breathing exercises, graduated
tasks starting with gentle brief onsets and offsets were recommended for a combined
warm-up (Miller, 1990). Next would be agility patterns and then tasks aimed at smooth
transition in vocal registers (Miller 1990). Vocal register exercises involved ascending
and descending glides, arpeggios and scales. Vocalization from top to bottom of the
singer's vocalizing range is a recommended practice for maintaining the same tone across
the entire range (Hylton, 1995). Additionally, when the singer sings the upward and
downward scales, glissandi (gliding across a wide range) and, arpeggios, the vocal folds
Univers
ity of
Mala
ya
50
would be extended. All exercises except glissandi aim at endurance and vocal flexibility
while at the same time enabling the voice to function without pressure (Gish et al., 2012;
Hylton, 1995). A maximal stretch of vocal folds is achieved by using two-octave pitch
glides on /i/ or /u/. Such exercises both separately and together work the muscles of the
cricothyroid and thyroarytenoids and obtain the basic frequency over the first formant for
specific acoustic loads (Titze, 2001).
Another common feature to most vocal warm-ups and therapy regimens is the
concept of using the entire body as a musical instrument when singing. This is done
through exercises that enhance psychological and physical awareness of one’s breathing
and abdominal support. Most vocal coaches emphasized the importance of abdominal or
diaphragm muscle support, which helps the singer to have better control over the way
they breathe. Having more control over one’s breathing influences tone quality, phonation
range capabilities, dynamics and intensity of the voice, and especially sound projection
(Thorpe, Cala, Chapman & Davis, 2001). It is necessary for singing that breath be
contained “low,” to seem as though one is singing “without breath,” by not quickly
expending one’s air intake, as this allows the singer to take in more breath and also to
sustain breath throughout singing activities (DeCaro & Donato, 2006). In practising low
breath, the singer does not expend excess air and, when coupled with enhanced resonance
control, a stronger voice can be produced with less effort (Stemple et al., 2010). Due to
the necessity for breath support, it is then obvious that good posture is also a necessity for
optimal breathing and voice production as the abdominal muscles support the singing
voice (DeCaro & Donato, 2006). According to Mendes et al. (2003), effective vocal
training and warm-up activities include maintaining the correct posture, articulatory
precision and vocal function exercises, strengthening the abdominal muscles with
breathing exercises. Vocal warm-ups thus have a strong psychological component, as well
Univers
ity of
Mala
ya
51
as helping to facilitate relaxation and mental readiness for performing (McHenry et al.,
2009).
Singers and non-singers alike have found that vocal warm-ups positively enhance
the vocal mechanism physically and acoustically. Warm-up encourages improved breath
support, thereby allowing singers to become conscious of the low breath, as well as
singing “without breath” to avoid laryngeal tension to be able to better sustain passages
(Donato & DeCaro, 2006). Thorpe et al. (2001) studied five professional singers by
recording their singing with and without breath support while examining their breathing
patterns with sensors attached to their skin at the abdominal and chest area. Their results
showed that when singers used a supported voice, glottal quotient was lower. The glottal
quotient is defined as the part of the glottal cycle in which the glottis is openly divided by
the amount of time in which it is closed. The more open glottal quotient is typically a
negative quality and is usually seen as an indicator of breathiness and is often a symptom
of a voice disorder (Henrich, d’Alessandro, Doval & Castellengo, 2005). Thorpe et al.
(2001) also found that the singer's formant increased by around 3 kHz and that the average
flow rate decreased significantly, indicating that breaths terminated at higher volumes of
the lungs. Having breaths terminate at higher lung volumes was interpreted to show that
with a supported voice, the singer was able to have enough lung volume remaining to
prepare for the next phrase, allowing for less tension throughout their singing. Breath
support exercises are a staple of the singer’s warm-up for the reasons Thorpe et al. (2001)
demonstrated.
A survey in 1989 of nationally accredited singing teachers, reported 93 different
directions that singing teachers felt were most important for teaching breath support in
singing. The highest-ranked directions all dealt with proper body posture and abdominal
breathing (Griffin, Woo, Colton, Casper & Brewer, 1995). Griffin et al. (1995) also
Univers
ity of
Mala
ya
52
conducted a study in which they surveyed 8 classically trained singers about their
concepts of the characteristics of the supported singing voice and how it is produced. The
subjects in their study also confirmed that a singing voice with good breath support has
better tone quality and ease of management that an unsupported singing voice does not
have. They defined this better quality as being more resonant, clear, and manageable
across their ranges.
By this definition, a warm-up that requires singing with the adjustments required
for a supported voice can create a voice with optimal breath, resonance, and posture
support in any individual who makes those adjustments. According to Elliot, Sundberg,
and Gramming (1995), vocal warm-up has been shown to decrease muscle tension and
enhance overall laryngeal ability. This reduction in laryngeal tension, the enhancement
of breath support, and the increase ‘forward’ vocal resonance should support the non-
singer’s voice and also aid in the prevention of the development of voice disorders
(Vurma & Ross, 2003).
The warm-up exercise, according to Miller (1990), should end with fast arpeggios
and scale covering the entire vocal range. Singing arpeggiated passages or rapid-scale
facilitates lower jaw and throat muscle relaxation. It is a sign of excessive vocal stress
when a singer finds it difficult to sing rapid scales or arpeggios. Titze (2001) proposes to
perform staccato arpeggios to establish a dominant mode of vocal fold vibration. It also
exercises the muscles of the adductors and abductors in tandem with the tensor muscles
during the shift in pitch (Titze, 2001).
The common semi-occluded vocal tract warm-up component is humming or
sustained nasal consonants (Titze, 2001 & Gish et al., 2012). To develop a clean on-
gliding and off-gliding in sung vowels, vocal exercises aimed at resonance balancing are
important (Miller, 1986). Resonance balancing refers to the balanced combination
Univers
ity of
Mala
ya
53
between nasal and oral cavity resonances of the vowels, after any of the nasal consonants,
/m/,/n/ or /al/ (Miller, 1986). In the vocal warm-up, the application of nasal consonants
and vocal consonants is designed to enhance overall vocal quality and to ensure maximum
vocal economy through resonance (Verdolini, Druker, Palmer, & Samawi, 1998). The
maximized ratio between voice output (dB) and intraglottal impact stress (kPa) under
constant subglottal pressure and frequency conditions is the vocal economy, or “Ev-max”
(Verdolini et al., 1998). When the intraglottal pressure is decreased, voice output is
optimally maximized. In theory, nasal consonants & vowels also facilitate a sustained
oscillation of the vocal folds with an optimal glottal width (Milbrath & Solomon, 2003).
A reduction in PTP suggests phonatory ease (Titze, 1988) and a reduced pre-phonatory
glottal width is correlated with a decrease in PTP (Lucero, 1998). One of the voice
building technique is humming. It enables the singer to create a clear tone and a flow
phonation with minimal effort (Gregg, 1996). It also “projects the voice to the vocal
mask” (Shear, 2008). Some people warn against the frequent use of nasal consonants in
the music community. Nasal consonants can lead to undesired muscular habits such as
tongue stiffening, larynx rising, and soft palates to be reduced chronically (Gregg, 1996;
Nix, 1999). The lip trill, a semi-occlusive exercise, is a proposed alternative warm-
up phoneme (Nix, 1999).
Univers
ity of
Mala
ya
54
2.12 Summary
The aforementioned research has shown that vocal warm-up can make significant
improvements in the singing voice of singers. Vocal warm-up might also enhance the
singing voice of non-singers, yet little research exists in this area (Vinturri et al., 2001).
Non-singers often are used as the control to demonstrate the benefits and differences
found in the singing population, but little information exists as to the effect of a singing
warm-up on non-singers alone (Vinturri et al., 2001). Many of the elements of singing
involve adjustments to the vocal mechanism, lowering the larynx, adjusting the
articulators, and projecting resonance to a more ‘forward’ placement (DeCaro & Donato,
2006; Sundberg, 1974). Because describable physical adjustments are occurring during
singing, the non-singer could likely learn these concepts and adjustments to enhance their
singing voice.
Common warm-up exercises and their underlying physiological contributions to
vocal tract tuning were identified. Such routines and exercises differ in order and duration
among voice teachers and vocalists but the basic concept of warm-up within the singing
community is consistent. Due to the individual complex properties of the different tissue
layers of the vocal folds, attempts to quantify changes during phonation threshold
pressure (PTP) warm-up have produced mixed results. Nevertheless, it is also unclear
whether the larynx's intrinsic musculature, when warmed-up or subsequently exercised,
functions as the skeletal muscle. Changes in vocal tract configuration caused by a vocal
warm-up may improve the voice's resonant characteristics. Semi-occlusive vocal warm-
ups could make it easier, more efficient and more economical for the singer to
phonate after the exercise (Cielo et al., 2013; Kapsner-Smith et al., 2015). The crucial
thing is to understand how two different types of warm-up affect the vocal tract structure
and the corresponding acoustic measures of a trained voice: traditional vocal warm-up
Univers
ity of
Mala
ya
55
and semi-occluded vocal tract warm-up (Duke et al., 2015). Such knowledge in future
will serve as an important vocal training guide and in the management of voice therapy.
Standard acoustic measures for singing voice quality include perturbation
measures of shimmer and jitter, noise-to-harmonics ratio (NHR), Singing Power Ratio
(SPR), and relative accuracy of production (Amir et al., 2005). Such measures have been
used to distinguish the speaking voice from the singing voice, where the singing voice
has been found to have lower shimmer and NHR values as compared to the speaking
voice (Lundy et al., 2000). It stands to reason that trained singers would have lower jitter
values and more accurate pitch matching, as accurate tone production is a distinguishable
feature of singers versus non-singers (Murry, 1990; Amir et al., 2005).
In summary, the benefit of singing vocal warm-ups has been documented, and
non-singers could apply the techniques used by trained singers to enhance their singing
voices. However, few studies documented the impact of the singing warm-up on the voice
quality of the non-singer. Therefore, in this study, the acoustic measures of the untrained
female singer’s singing voice will be documented before and after a singing warm-up.
Individuals who have not had any previous vocal training will participate in a singing
vocal warm-up consisting of breathe support, resonant singing, and vocal tract shaping
exercises to stretch their range and enhance their voice and body awareness. It is
hypothesized that, like trained singers, untrained singers will also benefit positively from
a singing warm-up. A lack of research exists regarding the singing voices of untrained
singers and this study seeks to determine whether an untrained-singer can benefit from a
vocal singing warm-up in the same way as a singer.
Univers
ity of
Mala
ya
56
CHAPTER 3
RESEARCH METHODOLOGY
3.1 Research Design
This research aims to obtain information and draw conclusions about the impact
of vocal warm-up on the voice quality of untrained female singers. This chapter provides
information on the sample of the study; research design; instruments for data
measurement and collection; pilot study; and data analysis procedures. This is quasi-
experimental research conducted in a one-group pretest-posttest design. The dependent
variable is assessed once before and once after the treatment is applied (Sage, 2019).
Figure 2: One group Pretest - Posttest design (Sage, 2019)
3.2 Subjects
40 females between the ages of 21 and 35, who enjoyed singing as a hobby but
never had any previous vocal training were included in this study. The reason for choosing
untrained singers in this study was that in such subjects the vocal warm-up had a greater
effect on the voice than in professional singers who were constantly warmed up because
of the regular use of their professional voice. All participants were in good health and
have no history of voice or hearing problems. On the days of the experiment, the
participants were instructed to maintain their normal dietary and sleep patterns, and they
were told not to sing before the experiment. All participants were instructed to fill in the
One group Pre-test, Post-test Pre-test Treatment Post-test
O X O
Univers
ity of
Mala
ya
57
participant questionnaire (Appendix A). 40 participants were divided into eight groups
(five participants per group). The experiment for the eight groups were carried out in eight
different days. Research ethics application was filed in and was approved by The
University of Malaya Research Ethics Committee (UMREC) (Reference Number:
UM.TNC2/UMREC – 812).
3.3 Recording procedure and instrumentation
According to Richard (2004), a comfortable amount of warm-up time for a
beginner is between 10 and 20 minutes before feeling fatigued. Richard (2004) stated that
it may take 30 minutes for an advanced singer to touch all technical areas. Therefore, a
20-minute vocal warm-up has been chosen for the untrained female singers in this study.
Subjects were recorded before and immediately after a 20-minute vocal warm-up. To
ensure the validity of the measurements, participants were told not to sing or warm up
their voices before the recording. Each subject was recorded individually in a quiet room
while sustaining the vowels /a/, /o/ and /i/ in two different pitches: Low- A3 (220.0 Hz)
and High- C5 (523.2 Hz). The low pitch was produced in the chest register and the high
note in the head register. The singer was guided by each reference tone presented by the
piano recording, and the singer was instructed to sustain the produced vowels (target
tones) as accurate as possible for 5 seconds. The test was recorded twice based on the
above criteria, taking into account the best voice production. The signal was recorded
through a Neumann TLM102 microphone situated approximately 30cm from the
participant’s mouth, using a digital recording program, Studio One 4 Professional with a
frequency response of 44100 Hz with a 16-bit sound card and high-quality speakers. The
signal was pre-amplified with an Art Pro MPA-II amplifier.
Univers
ity of
Mala
ya
58
The subjects were instructed to warm-up their voices following a specific protocol
created by the author (Appendix B) after the first recording (“pre-warm-up” condition).
The warm-up consists of phonation and positioning using a collection of syllables sung
at different levels, registers and amplitude in different pitches. It also consists of breathing
exercises, scales, and triads, range extension exercises, and body posture and relaxation
tasks.
These exercises were based on activities established by several sources for the
implementation of the vocal warm-up procedure: Vocal warm-up for child choristers
(Falcão, Masson, Oliveira & Behlau, 2014), Vocal warm-up and fatigue (Milbrath &
Solomon, 2003), Vocal warm-up and cool-down: a systematic review (Ribeiro, Frigo,
Bastilha & Cielo, 2016). The exercises consist of general and specific vocal warm-ups as
detailed in Appendix B. Warm-ups started with general exercises that consist of stretching
and respiratory exercises, where participants have to follow presented instructions before
phonation. Breathing exercises consist of low focused abdominal breathing as well as
further stretching of the arms over the head and down to expand and stretch the ribcage
(Miller, 1990). The goal of these exercises was to improve the preparation and
mental concentration of the musical activity and to provide guidance for airflows and air
output controls. These are important conditions for singing and producing a richer sound.
Specific exercise such as humming, aimed at mobilizing and relaxing the mucosa
was then performed to improve glottal closure (Andrade, Cielo, Schwarz & Ribeiro,
2016). Exercises of bilabial nasal sound emission /m/ for a full second before moving into
the /a/ were performed for the vocal projection (Falcão, Masson, Oliveira & Behlau,
2014). To have a greater vocal extension and enhance sound modulation, ascending and
descending exercises were carried out (Gish et al., 2012; Hylton, 1995). Subjects were
Univers
ity of
Mala
ya
59
instructed to sing arpeggios scales of single-syllable words while focusing on breath and
posture support, as they extend up and down their phonation range.
The experimenter was present during the warm-ups to assist and coach subjects
and to ensure optimal understanding of the exercise. To minimize the effect of the time
of day on the participant's voice quality, all recording sessions took place in the evening,
performing the same tasks before and after the warm-up.
3.4 Acoustic Analysis
Each recorded data was analyzed using Praat, a software program that provides a
quantitative acoustic analysis of voice quality. For these analyses, only the central 3
seconds of the vowels /a/, /o/ and /i/ were taken into account. The initial attack phase and
the final release phase were avoided when major signal variations occur. The acoustic
parameters measured for each vowel were jitter, shimmer and harmonics-to-noise ratio
(HNR).
3.5 Statistical Analysis
The collected data were analyzed by the basic descriptive statistics, namely the
mean and standard deviation of the quantitative variables for each measurement period.
The normality distribution hypothesis was verified by the Shapiro-Wilk test for the
variables Jitter (per cent), Shimmer (per cent), and HNR (dB). All the variables were
normally distributed. A Paired-Samples T-test was performed independently for each of
three acoustic parameters. The statistical analysis was performed using the Statistical
Product and Service Solutions (SPSS) Version 25 to compare the differences found before
and after a vocal warm-up for the acoustical analyses and the level of significance was
set at 5% (p < 0.05).
Univers
ity of
Mala
ya
60
3.6 Pilot Study
To establish test reliability for the acoustic parameters that were used in this study,
a pilot study was conducted from September to October in the year 2019. Ten untrained
female singers between the ages of 21 and 35 were included in this pilot study. The vocal
acoustic analysis was performed using Praat Software.
3.7 Reliability
Reliability was established by using the Praat program to gather the data.
According to the Praat, “It is the most complete program available (it contains much more
than could be discussed here), or because it is distributed for free, but also because it
comes with the ®nest algorithms. The pitch analysis algorithm is the most accurate in the
world; the articulatory synthesis is the only one that can handle dynamic length changes
(ejectives), non-glottal myo-elastics (trills), and sucking effects (clicks, implosives); and
the gradual learning algorithm is the only linguistically-oriented learning algorithm that
can handle free variation” (Paul & Vincent, 2001). Besides, the microphone attached to
the Praat was placed at an appropriate distance from the participant during measurements,
and the distance was consistently maintained throughout subsequent sessions. Also,
background noise was limited by ensuring the door to the treatment room was closed for
all measurements.
Univers
ity of
Mala
ya
61
3.8 Data Analysis Procedure
The data collected were analyzed using quantitative measures. The data were
evaluated using Statistical Product and Service Solutions (SPSS) Version 25, utilizing
descriptive statistics and a Paired-Samples T-Test.
3.8.1 Research Question One
Does vocal warm-up significantly alter the acoustic parameter – Jitter - of the untrained
singing voice?
Mean values and standard deviations of the Jitter parameter were computed for all
of the three vowels /a/, /o/ and /i/ in both the Two Pitch-Conditions, Low-A3 (220.0 Hz)
and High-C5 (523.2 Hz), pre and post vocal warm-up. Paired-Samples T-Test was
conducted to explore the relationships between the independent variables and dependent
variables.
3.8.2 Research Question Two
Does vocal warm-up significantly alter the acoustic parameter – Shimmer - of the
untrained singing voice?
Mean values and standard deviations of the Shimmer parameter were computed
for all of the three vowels /a/, /o/ and /i/ in both the Two Pitch-Conditions, Low-A3 (220.0
Hz) and High-C5 (523.2 Hz), pre and post vocal warm-up. Paired-Samples T-Test was
conducted to explore the relationships between the independent variables and dependent
variables.
Univers
ity of
Mala
ya
62
3.8.3 Research Question Three
Does vocal warm-up significantly alter the acoustic parameter - Harmonics-to-noise
ratio (HNR) - of the untrained singing voice?
Mean Harmonics-to-noise ratio parameter values and standard deviations were
determined for all three vowels/a/, /o/ and/i/ in both Two Pitch-Conditions, Low-A3
(220.0 Hz) and High-C5 (523.2 Hz), pre and post vocal warm-up. Paired-Samples T-Test
was conducted to explore the relationships between the independent variables and
dependent variables.
3.8.4 Research Question Four
Do acoustic parameters (Jitter, Shimmer and Harmonics-to-noise ratio (HNR)) differ
with vowel type (/a/, /o/, /i/) at different pitch level – Low predetermined tone
(frequencies) - A3 (220.0 Hz)?
In low pitch-condition, A3 (220.0 Hz), pre and post vocal warm-up, the mean
values and standard deviations of Jitter, Shimmer and Harmonics-to-Noise ratio
parameters were measured for each of the vowels. Paired-Samples T-Test was conducted
to explore the relationships between the independent variables and dependent variables.
Univers
ity of
Mala
ya
63
3.8.5 Research Question Five
Do acoustic parameters (Jitter, Shimmer and Harmonics-to-noise ratio (HNR)) differ
with vowel type (/a/, /o/, /i/) at different pitch level – High predetermined tone
(frequencies) - C5 (523.2 Hz)?
Mean values and standard deviations of the Jitter, Shimmer and Harmonics-to-
noise ratio parameters were computed for each of the vowels /a/, /o/ and /i/ in the High
Pitch-Condition, C5 (523.2 Hz), pre and post vocal warm-up. Paired-Samples T-Test was
conducted to explore the relationships between the independent variables and dependent
variables.
Univers
ity of
Mala
ya
64
CHAPTER 4
RESULTS
This study investigated the effects of vocal warm-up on the voice quality of untrained
female singers in Malaysia. This chapter reports the findings and describes the
investigation.
1. Does vocal warm-up significantly alter the acoustic parameter - Jitter of the
untrained singing voice?
2. Does vocal warm-up significantly alter the acoustic parameter - Shimmer of the
untrained singing voice?
3. Does vocal warm-up significantly alter the acoustic parameter - Harmonics-to-
noise ratio (HNR) of the untrained singing voice?
4. Do acoustic parameters (Jitter, Shimmer and Harmonics-to-noise ratio (HNR))
differ with vowel type (/a/, /o/, /i/) at different pitch level – Low predetermined tone
(frequencies) - A3 (220.0 Hz)?
5. Do acoustic parameters (Jitter, Shimmer and Harmonics-to-noise ratio (HNR))
differ with vowel type (/a/, /o/, /i/) at different pitch level – High predetermined tone
(frequencies) - C5 (523.2 Hz)?
Data collected from the study were analysed using Praat Software and Paired-Samples T-
Test using the Statistical Package for the Social Science (SPSS) Version 25. In summary,
the findings indicated that jitter and shimmer significantly decrease while the harmonics-
to-noise ratio (HNR) significantly increase after a 20-minute structured vocal warm-up.
The Jitter and Shimmer decrease indicates that there is an improvement in vocal
Univers
ity of
Mala
ya
65
roughness and vocal breathiness respectively. The HNR increase indicates that there is an
improvement in vocal hoarseness.
Figure 3: Example of Vocal Acoustic Analysis by Praat
Univers
ity of
Mala
ya
66
Figure 4: Jitter, Shimmer and HNR parameters determined using Praat
Univers
ity of
Mala
ya
67
4.1 Research Question One
Does vocal warm-up significantly alter the acoustic parameter - Jitter of the untrained
singing voice?
The results of the statistical test showed a significant difference of the mean values
of the acoustic parameter – Jitter, measure before and after the vocal warm-up (Table 1)
which could be explained by the comparison tests.
Table 1 Mean Values and Standard Deviations (in Parentheses) of Jitter parameter for all the Vowels /a/, /o/ and /i/ in the Two Pitch-Conditions, Low- A3 (220.0 Hz) and High- C5 (523.2 Hz), before and after Vocal Warm-Up
Subject Jitter (%)
Before 0.31 (0.11) After 0.22 (0.07)
Table 2 Paired T-test Results of the Jitter parameter for all the Vowels /a/, /o/ and /i/ in the Two Pitch-Conditions, Low- A3 (220.0 Hz) and High- C5 (523.2 Hz), before and after Vocal Warm-Up
Paired Samples Test
Paired Differences Mean SD SE t df Sig (2-tailed)
Pair 1 Pretest - Posttest 0.09 0.09 0.006 15.09 239 0.00
Univers
ity of
Mala
ya
68
Table 1 demonstrated the mean values and standard deviations (in parentheses) of
Jitter parameter for all the Vowels /a/, /o/ and /i/ in the Two Pitch-Conditions, Low-A3
(220.0 Hz) and High-C5 (523.2 Hz), pre and post vocal warm-up. Table 2 demonstrated
the Paired T-test results of the Jitter parameter for all the Vowels /a/, /o/ and /i/ in the Two
Pitch-Conditions, Low-A3 (220.0 Hz) and High-C5 (523.2 Hz), before and after vocal
warm-up. The overall Jitter parameter, which was determined after warm-up, was less
(p<0.00) than the value before warm-up, for all the vowels in the two pitch conditions
(Low-A3 (220.0 Hz) and High-C5 (523.2 Hz). The value decreased from 0.31% to 0.22%.
A paired-sample t-test was performed to compare Jitter parameter Vowels /a/, /o/ and /i/
pre-warm-up and post-warm-up of both low (A3) and high pitch (C5) conditions. There
was a significant difference in the pre warm-up (M=0.31, SD=0.11) and post warm-up
(M=0.22, SD=0.07) conditions; t(239)=15.09, p=0.00. The null hypothesis of equal Jitter
pre and post-warm-up means was rejected, since p<0.05. Thus, the post-warm-up Jitter
mean was statistically significantly lower than the pre-warm-up Jitter mean.
Univers
ity of
Mala
ya
69
4.2 Research Question Two
Does vocal warm-up significantly alter the acoustic parameter - Shimmer of the untrained
singing voice?
The results of the statistical test showed a significant difference of the mean values
of the acoustic parameter – Shimmer, measure before and after the vocal warm-up (Table
3) which could be explained by the comparison tests.
Table 3
Mean Values and Standard Deviations (in Parentheses) of Shimmer parameter for all the Vowels /a/, /o/ and /i/ in the Two Pitch-Conditions, Low-A3 (220.0 Hz) and High-C5 (523.2 Hz), before and after Vocal Warm-Up
Subject Shimmer (%)
Before 4.76 (3.30) After 2.93 (1.66)
Table 4 Paired T-test Results of the Shimmer parameter for all the Vowels /a/, /o/ and /i/ in the Two Pitch-Conditions, Low-A3 (220.0 Hz) and High-C5 (523.2 Hz), before and after Vocal Warm-Up
Paired Samples Test
Paired Differences Mean SD SE t df Sig (2-tailed)
Pair 1 Pretest - Posttest 1.83 2.27 0.15 12.52 239 0.00
Univers
ity of
Mala
ya
70
Table 3 demonstrated the mean values and standard deviations (in parentheses) of
Shimmer parameter for all the Vowels /a/, /o/ and /i/ in the Two Pitch-Conditions, Low-
A3 (220.0 Hz) and High-C5 (523.2 Hz), pre and post vocal warm-up. Table 4
demonstrated the Paired T-test results of the Shimmer parameter for all the Vowels /a/,
/o/ and /i/ in the Two Pitch-Conditions, Low-A3 (220.0 Hz) and High-C5 (523.2 Hz),
before and after vocal warm-up. The overall Shimmer parameter was less than (p<0.00)
the value before the warm-up for all three vowels in the Two Pitch Conditions, the Low-
A3 (220.0 Hz) and the High C5 (523.2 Hz). The figure decreased from 4.76% to 2.93%.
A paired-sample t-test was performed to compare Shimmer parameter Vowels /a/, /o/ and
/i/ pre-warm-up and post-warm-up low (A3) and high pitch (C5) conditions. There was a
significant difference in the pre warm-up (M=4.76, SD=3.30) and post warm-up (M=2.93,
SD=1.66) conditions; t(239)=12.52, p=0.00. The null hypothesis of equal Shimmer pre
and post-warm-up means was rejected, since p<0.05. Thus, the post-warm-up Shimmer
mean was statistically significantly lower than the pre-warm-up Shimmer mean.
Univers
ity of
Mala
ya
71
4.3 Research Question Three
Does vocal warm-up significantly alter the acoustic parameter - Harmonics-to-noise ratio
(HNR) of the untrained singing voice?
The results of the statistical test showed a significant difference of the mean values
of the acoustic parameter – Harmonics-to-noise ratio (HNR), measure before and after
the vocal warm-up (Table 5) which could be explained by the comparison tests.
Table 5
Mean Values and Standard Deviations (in Parentheses) of Harmonics-to-noise ratio (HNR) parameter for all the Vowels /a/, /o/ and /i/ in the Two Pitch-Conditions, Low-A3 (220.0 Hz) and High-C5 (523.2 Hz), before and after Vocal Warm-Up
Subject HNR (%) Before 21.20 (5.02) After 24.54 (4.56)
Table 6 Paired T-test Results of the Harmonics-to-noise ratio (HNR) parameter for all the Vowels /a/, /o/ and /i/ in the Two Pitch-Conditions, Low-A3 (220.0 Hz) and High-C5 (523.2 Hz), before and after Vocal Warm-Up
Paired Samples Test
Paired Differences Mean SD SE t df Sig (2-tailed)
Pair 1 Pretest - Posttest -3.33 3.20 0.21 -16.14 239 0.00
Univers
ity of
Mala
ya
72
Table 5 demonstrated the mean values and standard deviations (in parentheses) of
Harmonics-to-noise ratio (HNR) parameter for all the Vowels /a/, /o/ and /i/ in the Two
Pitch-Conditions, Low-A3 (220.0 Hz) and High-C5 (523.2 Hz), pre and post vocal warm-
up. Table 6 demonstrated the Paired T-test results of the HNR parameter for all the
Vowels /a/, /o/ and /i/ in the Two Pitch-Conditions, Low-A3 (220.0 Hz) and High-C5
(523.2 Hz), before and after Vocal Warm-Up. The average HNR parameter was higher
after the warm-up (p<0.00) than the value before warm-up for all the three vowels in the
two pitch conditions, Low-A3 (220.0 Hz) and high-C5 (523.1 Hz). The value has
increased from 21.20dB to 24.54dB. A paired-sample t-test was performed to compare
HNR parameter Vowels /a/, /o/ and /i/pre-warm-up and post-warm-up for both low (A3)
and high pitch (C5) conditions. There was a significant difference in the pre warm-up
(M=21.20, SD=5.02) and post warm-up (M=24.54, SD=4.56) conditions; t(239)=-14.14,
p=0.00. The null hypothesis of equal HNR pre and post-warm-up means was rejected,
since p<0.05. Thus, the post-warm-up HNR mean was statistically significantly higher
than the pre-warm-up HNR mean.
Univers
ity of
Mala
ya
73
4.4 Research Question Four
Do acoustic parameters (Jitter, Shimmer and Harmonics-to-noise ratio (HNR)) differ with
vowel type (/a/, /o/, /i/) at different pitch level – Low predetermined tone (frequencies) -
A3 (220.0 Hz)?
4.4.1 Jitter
Table 7 Mean Values and Standard Deviations (in Parentheses) of Jitter for the Vowels /a/, /o/ and /i/ in the Low Pitch-Condition, A3 (220.0 Hz), before and after Vocal Warm-Up
Vowel /a/ Vowel /o/ Vowel /i/ Variable Warmup A3 A3 A3 Jitter Before 0.37 0.36 0.37 (0.07) (0.10) (0.09) After 0.25 0.25 0.25 (0.07) (0.09) (0.06)
Table 8
Paired T-test Results of the Jitter parameter classified by vowels
A3 Subject Jitter
a t=10.82 p=0.00 o t=6.29 p=0.00 i t=7.97 p=0.00
Univers
ity of
Mala
ya
74
After the vocal warm-up, the average vowel/a/ jitter parameter in low pitch
condition (A3) was lower (p<0.00) than the value before warm-up which decreased from
0.37 to 0.25 per cent. A paired-sample t-test was performed to compare Jitter parameter
Vowel /a/ pre-warm-up and post-warm-up low pitch (A3) conditions. There was a
significant difference in the pre warm-up (M=0.37, SD=0.07) and post warm-up (M=0.25,
SD=0.07) conditions; t(39)=10.82, p=0.00. The null hypothesis of equal Jitter pre and
post-warm-up means was rejected, since p<0.05. Thus, the post-warm-up Jitter mean was
statistically significantly lower than the pre-warm-up Jitter mean.
The average vowel /o/ Jitter parameter in low pitch condition (A3) calculated after
the vocal warm-up was lower (p<0.00) than the pre-warm-up value, which decreased
from 0.36 to 0.25 per cent. A paired-sample t-test was performed to compare Jitter
parameter Vowel /o/ pre-warm-up and post-warm-up low pitch (A3) conditions. There
was a significant difference in the pre warm-up (M=0.36, SD=0.10) and post warm-up
(M=0.25, SD=0.09) conditions; t(39)=6.29, p=0.00. The null hypothesis of equal Jitter
pre and post-warm-up means was rejected, since p<0.05. Thus, the post-warm-up Jitter
mean was statistically significantly lower than the pre-warm-up Jitter mean.
The average vowel /i/ Jitter parameter in low pitch condition (A3) calculated after
the vocal warm-up was lower (p<0.00) than the pre-warm-up value, decreasing from 0.37
to 0.25 per cent. A paired-sample t-test was performed to compare Jitter parameter Vowel
/i/ pre-warm-up and post-warm-up low pitch (A3) conditions. There was a significant
difference in the pre warm-up (M=0.37, SD=0.09) and post warm-up (M=0.25, SD=0.06)
conditions; t(39)=7.97, p=0.00. The null hypothesis of equal Jitter pre and post-warm-up
means was rejected, since p<0.05. Thus, the post-warm-up Jitter mean was statistically
significantly lower than the pre-warm-up Jitter mean.
Univers
ity of
Mala
ya
75
The difference between the three vowels was evident in the Jitter parameter. Table
7 indicated that the Jitter values are higher for the /a/ and/i/ vowels than the non-high
vowel /o/ before the vocal warm-up. The findings after vocal warm-up indicated that all
three vowels have the same Jitter values.
Figure 5 illustrated the average Jitter values measured pre and post vocal warm-
up and their standard deviations at the low pitch level (A3).
Figure 5: Group mean values and standard deviations for the Jitter parameter at Low pitch level (A3), pre and post vocal warm-up
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0.45
0.5
A3 A3 A3
Vowel /a/ Vowel /o/ Vowel /i/
Jitt
er (
%)
Pre Warm-Up Post Warm-Up
Univers
ity of
Mala
ya
76
4.4.2 Shimmer
Table 9 Mean Values and Standard Deviations (in Parentheses) of Shimmer for the Vowels /a/, /o/ and /i/ in the Low Pitch-Condition, A3 (220.0 Hz), before and after Vocal Warm-Up
Vowel /a/ Vowel /o/ Vowel /i/ Variable Warmup A3 A3 A3 Shimmer Before 5.61 6.30 8.97 (1.98) (2.47) (4.18) After 2.96 3.79 5.30 (0.90) (1.40) (2.36)
Table 10 Paired T-test Results of the Shimmer parameter classified by vowels
A3 Subject Shimmer
a t=9.04 p=0.00 o t=8.81 p=0.00 i t=6.70 p=0.00
Univers
ity of
Mala
ya
77
The average vowel /a/ Shimmer parameter in low pitch condition (A3) calculated
following the vocal warm-up was lower (p<0.00) than the value before the warm-up,
decreasing from 5.61 to 2.96 per cent. A paired-sample t-test was performed to compare
Shimmer parameter Vowel /a/ pre-warm-up and post-warm-up low pitch (A3) conditions.
There was a significant difference in the pre warm-up (M=5.61, SD=1.98) and post warm-
up (M=2.96, SD=0.90) conditions; t(39)=9.04, p=0.00. The null hypothesis of equal
Shimmer pre and post-warm-up means was rejected, since p<0.05. Thus, the post-warm-
up Shimmer mean was statistically significantly lower than the pre-warm-up Shimmer
mean.
The average vowel /o/ Shimmer parameter in low pitch condition (A3) calculated
following the vocal warm-up was lower (p<0.00) than the value before the warm-up,
decreasing from 6.30 to 3.79 per cent. A paired-sample t-test was performed to compare
Shimmer parameter Vowel /o/ pre-warm-up and post-warm-up low pitch (A3) conditions.
There was a significant difference in the pre warm-up (M=6.30, SD=2.47) and post warm-
up (M=3.79, SD=1.40) conditions; t(39)=8.81, p=0.00. The null hypothesis of equal
Shimmer pre and post-warm-up means was rejected, since p<0.05. Thus, the post-warm-
up Shimmer mean was statistically significantly lower than the pre-warm-up Shimmer
mean.
The average vowel /i/ Shimmer parameter in low pitch condition (A3) calculated
following the vocal warm-up was lower (p<0.00) than the value before the warm-up,
decreasing from 8.97 to 5.30 per cent. A paired-sample t-test was performed to compare
Shimmer parameter Vowel /i/ pre-warm-up and post-warm-up low pitch (A3) conditions.
There was a significant difference in the pre-warm-up (M=8.97, SD=4.18) and post-
warm-up (M=5.30, SD=2.36) conditions; t(39)=6.83, p=0.000. The null hypothesis of
equal Shimmer pre and post-warm-up means was rejected, since p<0.05. Thus, the post-
Univers
ity of
Mala
ya
78
warm-up Shimmer mean was statistically significantly lower than the pre-warm-up
Shimmer mean.
The difference between the three vowels was evident in the Shimmer parameter.
Table 9 indicated that the Shimmer values are higher for the vowel /i/ than the non-high
vowels /a/ and /o/ pre and post vocal warm-up.
Figure 6 illustrated the average Shimmer values measured pre and post vocal
warm-up and their standard deviations at the low pitch level (A3).
Figure 6: Group mean values and standard deviations for the Shimmer parameter at Low pitch level (A3), pre and post vocal warm-up
0
2
4
6
8
10
12
14
A3 A3 A3
Vowel /a/ Vowel /o/ Vowel /i/
Shim
mer
(%
)
Pre Warm-Up Post Warm-Up
Univers
ity of
Mala
ya
79
4.4.3 Harmonics-to-noise ratio
Table 11 Mean Values and Standard Deviations (in Parentheses) of Harmonics-to-noise ratio for the Vowels /a/, /o/ and /i/ in the Low Pitch-Condition, A3 (220.0 Hz), before and after Vocal Warm-Up
Vowel /a/ Vowel /o/ Vowel /i/ Variable Warmup A3 A3 A3 HNR Before 16.36 18.68 17.57 (3.15) (3.35) (3.27) After 21.55 22.78 20.47 (2.96) (3.09) (3.14)
Table 12
Paired T-test Results of the Shimmer parameter classified by vowels
A3 Subject HNR
a t= -11.33 p= 0.00 o t= -9.48 p= 0.00 i t= -6.24 p= 0.00
Univers
ity of
Mala
ya
80
The average Harmonics-to-Noise ratio (HNR) of the low-pitch condition (A3)
vowel /a/ measured in warm-up conditions was higher (p<0.00) than before the warm-up
session, increasing from 16.36 to 21.55 dB. A paired-sample t-test was performed to
compare HNR parameter Vowel /a/ pre-warm-up and post-warm-up low pitch (A3)
conditions. There was a significant difference in the pre warm-up (M=16.36, SD=3.15)
and post warm-up (M=21.55, SD=2.96) conditions; t(39)=-11.33, p=0.00. The null
hypothesis of equal HNR pre and post-warm-up means was rejected, since p<0.05. Thus,
the post-warm-up HNR mean was statistically significantly higher than the pre-warm-up
HNR mean.
The average Harmonics-to-Noise ratio (HNR) of the low-pitch condition (A3)
vowel /o/ measured in warm-up conditions was higher (p<0.00) than before the warm-up
session, increasing from 18.68 to 22.78 dB. A paired-sample t-test was performed to
compare HNR parameter Vowel /o/ pre-warm-up and post-warm-up low pitch (A3)
conditions. There was a significant difference in the pre warm-up (M=16.68, SD=3.35)
and post warm-up (M=22.78, SD=3.09) conditions; t(39)=-9.48, p=0.00. The null
hypothesis of equal HNR pre and post-warm-up means was rejected, since p<0.05. Thus,
the post-warm-up HNR mean was statistically significantly higher than the pre-warm-up
HNR mean.
The average Harmonics-to-Noise ratio (HNR) of the low-pitch condition (A3)
vowel /i/ measured in warm-up conditions was higher (p<0.00) than before the warm-up
session, increasing from 17.57 to 20.47 dB. A paired-sample t-test was performed to
compare HNR parameter Vowel /i/ pre-warm-up and post-warm-up low pitch (A3)
conditions. There was a significant difference in the pre warm-up (M=17.57, SD=3.27)
and post warm-up (M=20.47, SD=3.14) conditions; t(39)=-6.241, p=0.00. The null
hypothesis of equal HNR pre and post-warm-up means was rejected, since p<0.05. Thus,
Univers
ity of
Mala
ya
81
the post-warm-up HNR mean was statistically significantly higher than the pre-warm-up
HNR mean.
The difference between the three vowels was evident in the HNR parameter. Table
11 indicated that the HNR values are higher for the non-high vowel /o/ than the high
vowel /i/ before the vocal warm-up. The findings after vocal warm-up indicated that the
non-high vowels /a/ and /o/ had higher values of HNR than the high vowel /i/.
Figure 7 illustrated the average Harmonics-to-noise ratio (HNR) values measured
pre and post vocal warm-up and their standard deviations at a low pitch level (A3).
Figure 7: Group mean values and standard deviations for the Harmonics-to-noise ratio (HNR) parameter at Low pitch level (A3), pre and post vocal warm-up
0.00
5.00
10.00
15.00
20.00
25.00
30.00
A3 A3 A3
Vowel /a/ Vowel /o/ Vowel /i/
Har
mon
ics-
to-n
oise
ratio
(dB
)
Pre Warm-Up Post Warm-UpUnivers
ity of
Mala
ya
82
4.5 Research Question Five
Do acoustic parameters (Jitter, Shimmer and Harmonics-to-noise ratio (HNR)) differ with
vowel type (/a/, /o/, /i/) at different pitch level – High predetermined tone (frequencies) -
C5 (523.2 Hz)?
4.5.1 Jitter
Table 13 Mean Values and Standard Deviations (in Parentheses) of Jitter for the Vowels /a/, /o/ and /i/ in the High Pitch-Condition, C5 (523.2 Hz), before and after Vocal Warm-Up
Vowel /a/ Vowel /o/ Vowel /i/ Variable Warmup C5 C5 C5 Jitter Before 0.27 0.25 0.26 (0.11) (0.09) (0.07) After 0.21 0.18 0.21 (0.07) (0.07) (0.06)
Table 14 Paired T-test Results of the Jitter parameter classified by vowels
C5 Subject Jitter
a t=3.60 p=0.001 o t=5.97 p=0.00 i t=5.271 p=0.00
Univers
ity of
Mala
ya
83
The average vowel /a/ Jitter parameter in high pitch condition (C5) calculated
following the vocal warm-up was lower (p<0.001) than the value before the warm-up,
decreasing from 0.27 to 0.21 per cent. A paired-sample t-test was performed to compare
Jitter parameter Vowel /a/ pre-warm-up and post-warm-up high pitch (C5) conditions.
There was a significant difference in the pre warm-up (M=0.27, SD=0.11) and post warm-
up (M=0.21, SD=0.07) conditions; t (39) =3.60, p=0.001. The null hypothesis of equal
Jitter pre and post-warm-up means was rejected, since p<0.05. Thus, the post-warm-up
Jitter mean was statistically significantly lower than the pre-warm-up Jitter mean.
The average vowel /o/ Jitter parameter in high pitch condition (C5) calculated
following the vocal warm-up was lower (p<0.00) than the value before the warm-up,
decreasing from 0.25 to 0.18 per cent. A paired-sample t-test was performed to compare
Jitter parameter Vowel /o/ pre-warm-up and post-warm-up high pitch (C5) conditions.
There was a significant difference in the pre warm-up (M=0.25, SD=0.09) and post warm-
up (M=0.18, SD=0.07) conditions; t (39) =5.97, p=0.00. The null hypothesis of equal
Jitter pre and post-warm-up means was rejected, since p<0.05. Thus, the post-warm-up
Jitter mean was statistically significantly lower than the pre-warm-up Jitter mean.
The average vowel /i/ Jitter parameter in high pitch condition (C5) calculated
following the vocal warm-up was lower (p<0.00) than the value before the warm-up,
decreasing from 0.26 to 0.21 per cent. A paired-sample t-test was performed to compare
Jitter parameter Vowel /i/ pre-warm-up and post-warm-up high pitch (C5) conditions.
There was a significant difference in the pre warm-up (M=0.26, SD=0.07) and post warm-
up (M=0.21, SD=0.06) conditions; t (39) =5.271, p=0.00. The null hypothesis of equal
Jitter pre and post-warm-up means was rejected, since p<0.05. Thus, the post-warm-up
Jitter mean was statistically significantly lower than the pre-warm-up Jitter mean.
Univers
ity of
Mala
ya
84
The difference between the three vowels was evident in the Jitter parameter. Table
13 indicated that the Jitter values are higher for the non-high vowel /a/ and the high-vowel
/i/ than the non- high vowel /o/ before the vocal warm-up. The findings after vocal warm-
up indicated that the high vowel /i/ and the non-high vowels /a/ had higher values of Jitter
than the non-high vowel /o/.
Figure 8 illustrated the average Jitter values measured pre and post vocal warm-
up and their standard deviations at high pitch level (C5).
Figure 8: Group mean values and standard deviations for the Jitter parameter at High pitch level (C5), pre and post vocal warm-up
Univers
ity of
Mala
ya
85
4.5.2 Shimmer
Table 15 Mean Values and Standard Deviations (in Parentheses) of Shimmer for the Vowels /a/, /o/ and /i/ in the High Pitch-Condition, C5 (523.2 Hz), before and after Vocal Warm-Up
Vowel /a/ Vowel /o/ Vowel /i/ Variable Warmup C5 C5 C5 Shimmer Before 2.66 2.30 2.19 (0.88) (0.76) (0.69) After 2.16 1.70 1.68 (0.84) (0.53) (0.49)
Table 16 Paired T-test Results of the Shimmer parameter classified by vowels
C5 Subject Shimmer
a t=3.98 p=0.00 o t=4.75 p=0.00 i t=7.09 p=0.00
Univers
ity of
Mala
ya
86
The average vowel /a/ Shimmer parameter in high pitch condition (C5) calculated
following the vocal warm-up was lower (p<0.00) than the value before the warm-up,
decreasing from 2.66 to 2.16 per cent. A paired-sample t-test was performed to compare
Shimmer parameter Vowel /a/ pre-warm-up and post-warm-up high pitch (C5)
conditions. There was a significant difference in the pre warm-up (M=2.66, SD=0.88)
and post warm-up (M=2.16, SD=0.84) conditions; t(39)=3.98, p=0.00. The null
hypothesis of equal Shimmer pre and post-warm-up means was rejected, since p<0.05.
Thus, the post-warm-up Shimmer mean was statistically significantly lower than the pre-
warm-up Shimmer mean.
The average vowel /o/ Shimmer parameter in high pitch condition (C5) calculated
following the vocal warm-up was lower (p<0.00) than the value before the warm-up,
decreasing from 2.30 to 1.70 per cent. A paired-sample t-test was performed to compare
Shimmer parameter Vowel /o/ pre-warm-up and post-warm-up high pitch (C5)
conditions. There was a significant difference in the pre warm-up (M=2.30, SD=0.76)
and post warm-up (M=1.70, SD=0.53) conditions; t(39)=4.75, p=0.00. The null
hypothesis of equal Shimmer pre and post-warm-up means was rejected, since p<0.05.
Thus, the post-warm-up Shimmer mean was statistically significantly lower than the pre-
warm-up Shimmer mean.
The average vowel /i/ Shimmer parameter in high pitch condition (C5) calculated
following the vocal warm-up was lower (p<0.00) than the value before the warm-up,
decreasing from 2.19 to 1.68 per cent. A paired-sample t-test was performed to compare
Shimmer parameter Vowel /i/ pre-warm-up and post-warm-up high pitch (C5) conditions.
There was a significant difference in the pre warm-up (M=2.19, SD=0.69) and post warm-
up (M=1.68, SD=0.49) conditions; t(39)=7.09, p=0.00. The null hypothesis of equal
Shimmer pre and post-warm-up means was rejected, since p<0.05. Thus, the post-warm-
Univers
ity of
Mala
ya
87
up Shimmer mean was statistically significantly lower than the pre-warm-up Shimmer
mean. Figure 9 demonstrated the graphs of Shimmer mean values and their standard
deviation at high pitch level (C5) measured before and after vocal warm-up.
The difference between the three vowels was evident in the Shimmer parameter.
Table 15 indicated that the Shimmer values are higher for the non-high vowel /a/ than the
vowel /i/ and /o/ before the vocal warm-up. The findings after vocal warm-up
indicated that non-high vowels /a/ had higher values of Shimmer than the high vowel /i/
and non-high vowel /o/.
Figure 9 illustrated the average Shimmer values measured pre and post vocal
warm-up and their standard deviations at high pitch level (C5).
Figure 9: Group mean values and standard deviations for the Shimmer parameter at High pitch level (C5), pre and post vocal warm-up
Univers
ity of
Mala
ya
88
4.5.3 Harmonics-to-noise ratio
Table 17 Mean Values and Standard Deviations (in Parentheses) of Harmonics-to-noise ratio (HNR) for the Vowels /a/, /o/ and /i/ in the High Pitch-Condition, C5 (523.2 Hz), before and after Vocal Warm-Up
Vowel /a/ Vowel /o/ Vowel /i/ Variable Warmup C5 C5 C5 HNR Before 23.67 25.93 25.02 (3.10) (2.94) (3.97) After 26.83 28.98 26.64 (3.23) (3.67) (4.00)
Table 18 Paired T-test Results of the Harmonics-to-noise ratio (HNR) parameter classified by vowels
C5 Subject HNR
a t= -7.09 p= 0.00 o t= -6.10 p= 0.00 i t= -2.855 p= 0.007
Univers
ity of
Mala
ya
89
The average Harmonics-to-Noise ratio (HNR) of the high-pitch condition (C5)
vowel /a/ measured in warm-up conditions was higher (p<0.00) than before the warm-up
session, increasing from 23.67 to 26.83 dB. A paired-sample t-test was performed to
compare HNR parameter Vowel /a /pre-warm-up and post-warm-up high pitch (C5)
conditions. There was a significant difference in the pre warm-up (M=23.67, SD=3.10)
and post warm-up (M=26.83, SD=3.23) conditions; t(39)= -7.09, p=0.00. The null
hypothesis of equal HNR pre and post-warm-up means was rejected, since p<0.05. Thus,
the post-warm-up HNR mean was statistically significantly higher than the pre-warm-up
HNR mean.
The average Harmonics-to-Noise ratio (HNR) of the high-pitch condition (C5)
vowel /o/ measured in warm-up conditions was higher (p<0.00) than before the warm-up
session, increasing from 25.93 to 28.98 dB. A paired-sample t-test was performed to
compare HNR parameter Vowel /o/ pre-warm-up and post-warm-up high pitch (C5)
conditions. There was a significant difference in the pre warm-up (M=25.93, SD=2.94)
and post warm-up (M=28.98, SD=3.67) conditions; t(39)= -6.10, p=0.00. The null
hypothesis of equal HNR pre and post-warm-up means was rejected, since p<0.05. Thus,
the post-warm-up HNR mean was statistically significantly higher than the pre-warm-up
HNR mean.
The average Harmonics-to-Noise ratio (HNR) of the high-pitch condition (C5)
vowel /i/ measured in warm-up conditions was higher (p<0.007) than before the warm-
up session, increasing from 25.02 to 26.64 dB. A paired-sample t-test was performed to
compare HNR parameter Vowel /i/ pre-warm-up and post-warm-up high pitch (C5)
conditions. There was a significant difference in the pre warm-up (M=25.02, SD=3.97)
and post warm-up (M=26.64, SD=4.00) conditions; t(39)= -2.85, p=0.007. The null
hypothesis of equal HNR pre and post-warm-up means was rejected, since p<0.05. Thus,
Univers
ity of
Mala
ya
90
the post-warm-up HNR mean was statistically significantly higher than the pre-warm-up
HNR mean.
The difference between the three vowels was evident in the HNR parameter. Table
17 indicated that the HNR values are higher for the non-high vowel /o/ than the high
vowel /i/ and non-high vowel /a/ pre vocal warm-up. The findings after vocal warm-up
indicated that the non-high vowel /a/ had higher values of HNR than the high vowel /i/
and non-high vowel /o/.
Figure 10 illustrated the average Harmonics-to-noise ratio (HNR) values
measured pre and post vocal warm-up and their standard deviations at high pitch level
(C5).
Figure 10: Group mean values and standard deviations for the Harmonics-to-noise ratio (HNR) parameter at High pitch level (C5), pre and post vocal warm-up
Univers
ity of
Mala
ya
91
CHAPTER 5
DISCUSSIONS, CONCLUSIONS AND RECOMMENDATIONS
5.1 Summary
This research aimed to determine the effect of vocal warm-up on the voice quality
of 40 untrained female singers in Malaysia. In this study, vocal acoustic parameters were
measured pre and post-warm-up session, with an intervention of phonatory activity, and
significant effects on all acoustic parameters were identified.
This study reflects the impact of vocal warm-up on the three acoustic parameters:
jitter and shimmer decrease while the harmonics-to-noise ratio (HNR) increase after a 20-
minute structured vocal warm-up. The Jitter and Shimmer decrease indicates that there is
an improvement in vocal roughness and vocal breathiness respectively. The HNR
increase indicates that there is an improvement in vocal hoarseness. The benefit of singing
vocal warm-ups has been documented, and untrained singers could apply the techniques
used by trained singers to enhance their singing voices.
Univers
ity of
Mala
ya
92
5.2 Discussions of the Research Questions
This section will discuss each research question posed for this study. Each research
question will be discussed individually.
5.2.1 Research Question One
Does vocal warm-up significantly alter the acoustic parameter - Jitter of the untrained
singing voice?
Table 1 demonstrated a significant decrease in Jitter values after vocal warm-up.
The vocal folds display fundamental frequency and amplitude variations during sustained
vibration. The Jitter reflects the “frequency perturbation”. Jitter demonstrates the
variation in vocal fold vibration and is related to the perception of vocal roughness. The
Jitter increases in patients with vocal fold disorders. The decrease shown suggests that
the vocal warm-up influences the vibration cycle in the vocal fold and also decreases
perturbations in normophonic subjects. It also indicates that the vibration adjustment of
the vocal folds has improved, allowing for better use of the voice during the performance.
The lack of control of the vocal fold vibration mainly affects Jitter (De Felippe, Grillo &
Grechi, 2006; Teixeira, Oliveira, & Lopes, 2013; Mezzedimi et al., 2018).
Univers
ity of
Mala
ya
93
5.2.2 Research Question Two
Does vocal warm-up significantly alter the acoustic parameter - Shimmer of the untrained
singing voice?
Table 3 demonstrated a significant decrease in Shimmer values after vocal warm-
up. The “amplitude perturbation” is represented by Shimmer. The shimmer is related to
sound wave amplitude or vocal emission intensity. This is primarily caused by the
reduction of glottis resistance and mass lesions in the vocal folds, which are associated
with noise at emission and breathiness. The Shimmer also shows disturbances of short-
term amplitude that are often minimally present in normophonic subjects. The
relationship between subglottic pressure and glottis resistance is particularly affected by
Shimmer. The decrease we observed indicates that the vocal warm-up affected the
mechanisms underlying the phonatory effort, reducing the amplitude changes shown by
the shimmer (Teixeira, Oliveira, & Lopes, 2013; James & John, 2007; Mezzedimi et al.,
2018).
Univers
ity of
Mala
ya
94
5.2.3 Research Question Three
Does vocal warm-up significantly alter the acoustic parameter - Harmonics-to-noise ratio
(HNR) of the untrained singing voice?
Table 5 demonstrated a significant increase of HNR values after vocal warm-up.
The Harmonics-to-noise ratio (HNR) is defined as the quantity and harmonic energy of
the fundamental frequency (F0), divided by noise energy frequencies. HNR is related to
the perception of vocal roughness and breathiness. Disordered voices have a high level of
noise and low HNR. A decrease in HNR can indicate either increased additive disturbance
noise associated with impaired glottis closure (breathiness) or increased jitter (roughness)
(Teixeira, Oliveira, & Lopes, 2013; James & John, 2007). The increase in the results of
this study indicates that the amount of periodic vocal signals is higher after vocal warm-
up than before. Besides, this can also be correlated with a corresponding decline in
aperiodicity in the signal (Yumoto & Gould, 1982; De Felippe, Grillo & Grechi, 2006;
Teixeira, Oliveira, & Lopes, 2013; Mezzedimi et al., 2018).
Univers
ity of
Mala
ya
95
5.2.4 Research Question Four
Do acoustic parameters (Jitter, Shimmer and Harmonics-to-noise ratio (HNR)) differ with
vowel type (/a/, /o/, /i/) at different pitch level – Low predetermined tone (frequencies) -
A3 (220.0 Hz)?
Table 7 demonstrated a significant decrease of Jitter values for all three vowels
/a/, /o/ and /i/ at low pitch level (A3) which was produced in the chest register. As for the
vowel effect, according to Table 7, it can be seen that the non-high vowel /a/ and high
vowel /i/ had higher values of Jitter than the non-high vowel /o/ pre vocal warm-up. In
this study, pre vocal warm-up result is similar to Orlikoff’s study. In most acoustic studies
using multiple vowels, Orlikoff (1995) recorded low disturbance of high vowels and
higher disturbance of low vowels with several conflicting findings. However, in this
study, post vocal warm-up results showed that all three vowels have the same values of
Jitter. Therefore, the results could not be compared.
Table 9 demonstrated a significant decrease of Shimmer values for all three
vowels /a/, /o/ and /i/ at low pitch level (A3) which was produced in the chest register.
Shimmer results indicated that the high vowel /i/ had the highest values, intermediate for
low vowel /o/ and lowest for vowel /a/ for pre and post vocal warm-up. In a recent study,
the contrary results of the lowest shimmer for / i/, the intermediate for /a/ and the highest
for /u/ were reported by Akif et al. (2004). Vowels effect have less noticeable results in
terms of measuring jitter and shimmer. From these contradictory findings, it is clear that
the effects of vowels on the study of perturbation are inconclusive.
Table 11 demonstrated a significant increase of HNR values for all three vowels
/a/, /o/ and /i/ at low pitch level (A3) which was produced in the chest register. As for the
vowel effect, Table 11 demonstrated HNR highest values for vowel /o/, intermediate for
Univers
ity of
Mala
ya
96
vowel /i/ and lowest for vowel /a/ pre vocal warm-up. The SNR findings by MacCallum
et al. (2011) found similar vowel effects in females. SNR was highest for /i/, intermediate
for /u/, and lowest for /a/. In this study, post-warm-up HNR results showed the highest
values for vowel /o/, intermediate for the vowel /a/ and lowest for vowel /i/. The findings
of this study are inconsistent with the results of MacCallum et al. (2011).
5.2.5 Research Question Five
Do acoustic parameters (Jitter, Shimmer and Harmonics-to-noise ratio (HNR)) differ with
vowel type (/a/, /o/, /i/) at different pitch level – High predetermined tone (frequencies) -
C5 (523.2 Hz)?
Table 13 demonstrated a significant decrease of Jitter values for all three vowels
/a/, /o/ and /i/ at high pitch level (C5) which was produced in the head register. As for the
vowel effect, according to Table 13, it can be seen that the vowels /a/ and /i/ had higher
values of Jitter than the vowel /o/ pre vocal warm-up. Post vocal warm-up results showed
that the high vowel /i/ and non-high vowel /a/ had higher values of Jitter than the non-
high vowels /o/. Vowel /i/ and vowel /o/ had the same values, the results did not
distinguish high from low vowels. Therefore, the results were also not comparable.
Table 15 demonstrated a significant decrease of Shimmer values for all three
vowels /a/, /o/ and /i/ at high pitch level (C5) which was produced in the head register.
Shimmer results indicated that the high vowel /a/ had the highest values, intermediate for
the low vowel /i/ and lowest for vowel /o/ for pre and post vocal warm-up. The findings
in this study are in line with some researchers who have found Shimmer to be the lowest
for the vowel /u/, intermediate for vowel /i/, and highest for vowel /a/ (Horii, 1980;
Ramig, 1983; Sorensen & Horii, 1983).
Univers
ity of
Mala
ya
97
Table 17 demonstrated a significant increase of HNR values for all three vowels
/a/, /o/ and /i/ at high pitch level (C5) which was produced in the head register. HNR
results for pre vocal warm-up indicated that the non-high vowel /o/ had the highest values,
intermediate for the high vowel /i/ and lowest for vowel /a/. A few studies showed that
findings of the SNR analysis indicated that the production of low vowels, such as /a/, is
carried out in a lower vibration rate of vocal fold and induces greater signal noise levels
than high vowel production. This is known as the intrinsic pitch of vowels and is often
attributed to mechanical coupling resulting in anterior positioning of the hyoid bone and
forward tilting of the thyroid cartilage, creating an anterior pull on the vocal folds and
increasing fold tension for high vowels such as /i/ and /u/ (Higgins et al., 1998; Sapir,
1989). In this study, HNR results for post vocal warm-up indicated that the non-high
vowel /o/ had the highest values, intermediate for the non-high vowel /a/ and lowest for
the high vowel /i/. The results of this study are inconsistent with the results of MacCallum
et al. (2011). The low vowel /a/ had lesser harmonic activity in its signals, as indicated by
its lowest SNR values, and the slowest vibratory frequencies, as indicated by F0 values.
Conversely, the high vowels /i/ and /u/ had greater harmonic activity and higher vocal
fold vibratory frequencies and demonstrated less complexity.
Univers
ity of
Mala
ya
98
5.3 Implications of the Study
The significant reduction in jitter and shimmer and increase in HNR observed
after the vocal warm-up confirmed the positive effect and demonstrated the importance
of this practice. The findings in this study indicated that untrained singers could benefit
from vocal warm-up and it should be encouraged and not bypassed. Therefore,
professional singers or untrained singers should develop a vocal warm-up routine to
enhance their vocal qualities and prevent future voice problems. This study’s findings
could benefit voice teachers in implementing best practice in rehearsals. Future research
may investigate whether singing with various vowels and consonants has different effects
on the vocal mechanism and on the vocal acoustic parameters. It is also important to
compare the effects of vocal warm-up procedures of various duration and types in future
studies. Such research may establish the best procedure for vocal warm-up to improve
voice quality and prevent vocal injury in singers.
The findings of this study also seem to reinforce further suggestions of the
acoustic analysis as a valuable tool for assessing and measuring the impact of voice warm-
up on voice production (Teixeira & Fernandes, 2014; Sascha & Pascal, 2019; Bausar,
Bohlender, Mehta, 2018). Audio-visual feedback displays, such as those provided in the
Praat software version 6.1.03, might be useful tools in successfully developing a means
of warming up which is not only not damaging to the voice but beneficial in facilitating
an optimal vocal quality (Paul & Vincent, 2001). Using these objective measures of vocal
quality, one could develop a warm-up technique to establish an enhanced tone and
increased projection with minimal effort. One could also develop a warm-up technique to
establish a vocal quality with minimal vocal perturbation (i.e., pure tone quality) and
minimal variation in vocal perturbation (i.e., even vibrato), through the use of measures
of vocal perturbation (Sataloff, 2017; Sataloff, 2017).
Univers
ity of
Mala
ya
99
5.4 Conclusions
This study investigated the effects of vocal warm-in improving voice quality of
untrained female singers in two pitch-conditions, A3 (Chest register) & C5 (Head
register). The results demonstrated that after 20 minutes of vocal warm-up, all of the 40
participants’ voice qualities improved by decreasing Jitters and Shimmers, and increasing
Harmonics-to-noise ratio (HNR), which was consistent with other research studies (Amir
et al., 2005; Tay et al. 2012). The results show that acoustic analysis is a valuable tool for
assessing and measuring the impact of voice warm-up on voice production (Teixeira &
Fernandes, 2014; Sascha & Pascal, 2019; Bausar, Bohlender, Mehta, 2018). Such results
also support the importance of including different exercises in the warm-up routine which
are aimed not only at laryngeal muscles but also at breathing posture and relaxation. The
increase in amplitude perturbation measures after warm-up suggests that warm-up, in
addition to vocal fold control, also increases the regulation of the breathing
mechanism that has a significant role in amplitude variation.
The reduction in Jitter demonstrates that vocal warm-up affects the vibration cycle
of the vocal fold, which minimizes disturbances and enhances voice quality. The decline
in Shimmer indicates that it also affected the underlying mechanisms of the phonatory
effect, reducing the variations in amplitude. The Jitter and Shimmer decrease indicates
that there is an improvement in vocal roughness and vocal breathiness respectively. The
HNR increase indicates that there is an improvement in vocal hoarseness. Therefore, the
findings in this study indicated that untrained singers could benefit from vocal warm-up
and it should be encouraged and not bypassed (De Felippe, Grillo & Grechi, 2006;
Teixeira, Oliveira, & Lopes, 2013; Christian, 2007; Yumoto & Gould, 1982).
As for the vowel effect, the results indicated that vowel effects on perturbation
analysis are inconsistent among this and previous studies (Akif et al., 2004; MacCallum
Univers
ity of
Mala
ya
100
et al., 2011). Perturbation parameters have proved to be unpredictable as a result of vowel
effects in this analysis. After the vocal warm-up, all three vowels have the same values
of Jitter for the low pitch condition (A3). The results were not comparable. For per cent
Jitter in high pitch condition (C5), post-vocal warm-up findings indicated that the high
vowel /i/ and non-high vowel /a/ had higher values of Jitter than the non-high vowels /o/.
Vowel /i/ and vowel /o/ had the same values, the results did not distinguish high from low
vowels. Therefore, the results were also not comparable.
For per cent Shimmer in low pitch condition (A3), shimmer results indicated that
the high vowel /i/ had the highest values, intermediate for low vowel /o/ and lowest for
the vowel /a/ for pre and post vocal warm-up. The findings were not in line with the study
of Akif et al. (2004). Akif et al. (2004) found conflicting results including the opposite
findings of lowest shimmer for /i/, intermediate for /a/, and highest for /u/. For per cent
Shimmer in high pitch condition (C5), the results indicated that the high vowel /a/ had the
highest values, intermediate for the low vowel /i/ and lowest for vowel /o/ for pre and
post vocal warm-up. The results in this study are in line with some researchers who have
found Shimmer to be the lowest for the vowel /u/, intermediate for vowel /i/, and highest
for vowel /a/ (Horii, 1980; Ramig, 1983; Sorensen & Horii, 1983).
Overall, the results indicated that vowel effects on perturbation analysis are
inconsistent among this and previous studies. In recent years, it has been suggested that
the algorithms employed to calculate perturbation parameters may only be useful for
nearly periodic voice signals and may not reliably analyse strongly aperiodic signals
(Titze, 1995; Karnell, Chang, Smith & Hoffman, 1997). Besides, some recording and
analysis conditions including microphone type and placement (Titze & Winholtz, 1993),
analysis systems (Karnell, Scherer & Fischer, 1991; Bielamowicz, Kreiman, Gerratt,
Univers
ity of
Mala
ya
101
Dauer & Berke, 1996), and environmental noise (Carson, Ingrisano & Eggleston, 2003;
Deliyski, Shaw & Evans, 2005) have found to affect jitter and shimmer.
For Harmonics-to-noise ratio (HNR) in low pitch condition (A3), post-warm-up
HNR results showed the highest values for vowel /o/, intermediate for the vowel /a/ and
lowest for vowel /i/. The results of this study are inconsistent with the results of
MacCallum et al. (2011). He found that SNR was highest for /i/, intermediate for /u/, and
lowest for /a/. For Harmonics-to-noise ratio (HNR) in high pitch condition (C5), HNR
results for post vocal warm-up indicated that the non-high vowel /o/ had the highest
values, intermediate for the non-high vowel /a/ and lowest for the high vowel /i/. The
results of this study are inconsistent with the results of MacCallum et al. (2011). The low
/a/ vowel had less harmonic activity in its signals, as indicated by its lowest SNR values.
Conversely, the high vowels /i/ and /u/ had greater harmonic activity and higher vocal
fold vibratory frequencies and demonstrated less complexity.
This suggests that per cent jitter, per cent shimmer and harmonics-to-noise ratio
(HNR) are not useful parameters in reliably describing acoustic differences between
vowels. Vowel effects on acoustic perturbation analysis in other voice types, including
dysphonic voices, should be investigated to validate the acoustic effects of vowel
selection found in this research.
Univers
ity of
Mala
ya
102
5.5 Recommendations for future research
Singers from a wider range of geographical areas should be recruited in future
research to control regional variations. A control group was not used in this study, and to
ensure that results are related to the effect of the vocal warm-up, the use of a control group
in future studies could be beneficial. The impact of vocal warm-up on different genders
and ages in both men and women should be further explored for future research. More
vowel types should be explored to assess the effects on voice quality. Use of a larger
group of participants could also be conducted in future studies to enhance validity and
data significance. Besides, the present study used a single before and after measurements.
Future research using multiple before and after warm-up interventions could be useful.
For future studies, the effects of vowels on acoustic analysis could also be discussed.
In conclusion, through an acoustic analysis paradigm, the findings presented here
confirmed that vocal warm-up has significant evidence to increase the overall vocal
quality of untrained female singers. Further emphasis on the above studies would be of
great relevance and interest for singing teachers and their students.
Univers
ity of
Mala
ya
103
REFERENCES
Akif Kiliç, M., Öğüt, F., Dursun, G., Okur, E., Yildirim, I., & Midilli, R. (2004). The effects of vowels on voice perturbation measures. Journal of Voice, 18(3), 318–324. doi:10.1016/j.jvoice.2003.09.00
Amir, O., Amir, N., Michaeli, O. (2005). Evaluating the Influence of Warmup on Singing Voice Quality Using Acoustic Measures. Journal of Voice, 19, 252-260. Andrade, S. R., Cielo, C. A., Schwarz, K., & Ribeiro, V. V. (2016). Vocal therapy and
nasal sounds: effects on Hyperfunctional dysphonia. Revista CEFAC, 18(1), 263–272. doi:10.1590/1982-021620161810115
Baken, R. J., & Orlikoff, R. F. (1987). The effect of articulation on fundamental
frequency in singers and speakers. Journal of Voice, 1(1), 68–76. doi: 10.1016/s0892-1997(87)80027-9 Bausar, Bohlender, Mehta (2018) Acoustic Perturbation Measures Improve with
Increasing Vocal Intensity in Individuals With and Without Voice Disorders. Journal of Voice, 32(2), 162–168. doi:10.1016/j.jvoice.2017.04.008
Bele, I. V. (2006). The Speaker’s Formant. Journal of Voice, 20(4), 555–
578.doi:10.1016/j.jvoice.2005.07.001 Bielamowicz S, Kreiman J, Gerratt BR, Dauer MS, Berke GS (1996). Comparison of
voice analysis systems for perturbation measurement. J Speech Hear Res, 39, 126–134.
Blaylock, T. R. (1999). Effects of systematized vocal warm-up on voices with disorders
of various etiologies. Journal of Voice, 13(1), 43–50. doi: 10.1016/s0892-1997(99)80060-5
Brandon, C. A., Rosen, C., Georgelis, G., Horton, M. J., Mooney, M. P., & Sciote, J. J.
(2003). Muscle fibre type composition and effects of vocal fold immobilization on the two compartments of the human posterior cricoarytenoid: a case study of four patients. Journal of voice: official journal of the Voice Foundation, 17(1), 63–75. doi: 10.1016/s0892-1997(03)00027-4
Boone, D. (1988). Respiratory training in voice therapy. Journal of Voice, 2, 20-25.
Univers
ity of
Mala
ya
104
Brown, W. S., Jr., Morris, R. J., & Michel, J. F. (1990). Vocal jitter and fundamental frequency characteristics in aged, female professional singers. Journal of Voice, 4, 135-141.
Brown, W. S., Rothman, H. B., & Sapienza, C. M. (2000). Perceptual and acoustic study
of professionally trained versus untrained voices. Journal of Voice, 14(3), 301–309.doi:10.1016/s0892-1997 (00)80076-4
Carson CP, Ingrisano DRS, Eggleston KD (2003). The effect of noise on computer-aided
measures of voice: a comparison of CSpeechSP and the Multi-Dimensional Voice Program software using the CSL 4300B Module and Multi-Speech for Windows. J Voice 2003, 17, 12–20.
Chan, R. W., & Titze, I. R. (1998). Viscosities of Implantable Biomaterials in Vocal Fold
Augmentation Surgery. The Laryngoscope, 108(5), 725–731. doi: 10.1097/00005537-199805000-00019 Chernobelsky, S. I. (2007). The treatment and results of voice therapy amongst
professional classical singers with vocal fold nodules. Logopedics Phoniatrics Vocology, 32(4), 178–184. doi:10.1080/14015430600852043
Chhetri, D. K., Neubauer, J., Sofer, E., & Berry, D. A. (2014). Influence and interactions
of laryngeal adductors and cricothyroid muscles on fundamental frequency and glottal posture control. The Journal of the Acoustical Society of America, 135(4), 2052–2064. doi:10.1121/1.4865918
Christian Müller (2007). Speaker Classification I: Fundamentals, Features, and Methods.
Springer Science & Business Media, 2007. doi:
https://books.google.com.my/books?id=APi3_4AV3dkC&printsec=frontcover#v=onepage&q&f=false
Cielo, C. A., Lima, J. P. de M., Christmann, M. K., & Brum, R. (2013). Exercícios de
trato vocal semiocluído: revisão de literatura. Revista CEFAC, 15(6), 1679–1689. doi:10.1590/s1516-18462013005000041
Cleveland, T., Sundberg, J., Stone, R. (2001). Long-Term-Average Spectrum
Characteristics of Country Singers during Speaking and Singing. Journal of Voice, 15, 54-60.
Univers
ity of
Mala
ya
105
Cohen, S. M., Noordzij, J. P., Garrett, C. G., & Ossoff, R. H. (2008). Factors associated with perception of singing voice handicap. Otolaryngology-Head and Neck Surgery, 138(4), 430–434. doi:10.1016/j.otohns.2007.12.041
Cooper, D. S., & Titze, I. R. (1985). Generation and Dissipation of Heat in Vocal Fold
Tissue. Journal of Speech Language and Hearing Research, 28(2), 207. doi:10.1044/jshr.2802.207
Cooper, Donald & Partridge, LLoyd & Alipour, Fariborz. (1993). Muscle Energetics,
Vocal Efficiency, and Laryngeal Biomechanics. Singular Publishing Group, Inc DeCaro, J. B., & Donato, J. (2006). The way to sing: Secrets of bel canto. Teaneck, NJ: Feelgreat Publishing Co. DeFatta, R.A., & Sataloff, R.T. (2012). The Value of Vocal Warm-Up and Cool-Down
Exercises: Questions and Controversies. doi:
https://pdfs.semanticscholar.org/11ce/bafc001b12c6e24ba9adbac13b2931bae09f.pdf
Deguchi, S., Kawahara, Y., & Takahashi, S. (2011). Cooperative Regulation of Vocal
Fold Morphology and Stress by the Cricothyroid and Thyroarytenoid Muscles. Journal of Voice, 25(6), e255–e263. doi:10.1016/j.jvoice.2010.11.006
Deliyski DD, Shaw HS, Evans MK (2005). Adverse effects of environmental noise on
acoustic voice quality measurements. J Voice,19, 15–28. De Felippe, A. C. N., Grillo, M. H. M. M., & Grechi, T. H. (2006). Standardization of
acoustic measures for normal voice patterns. Brazilian Journal of Otorhinolaryngology, 72(5), 659–664.doi:10.1016/s1808-8694(15)31023-5
Duke, E., Plexico, L. W., Sandage, M. J., & Hoch, M. (2015). The Effect of Traditional
Singing Warm-Up Versus Semioccluded Vocal Tract Exercises on the Acoustic Parameters of Singing Voice. Journal of Voice, 29(6), 727–732.doi:10.1016/j.jvoice.2014.12.009
Ekholm, E., Papagiannis, G. C., & Chagnon, F. P. (1998). Relating objective
measurements to expert evaluation of voice quality in western classical singing: Critical perceptual parameters. Journal of Voice, 12(2), 182–196. doi: 10.1016/s0892-1997(98)80038-6
Univers
ity of
Mala
ya
106
Elliot, N., Sundberg, J., & Gramming, P. (1995). What happens during vocal warm-up? Journal of Voice, 9, 37-44.
Falcão, L. M. G., Masson, M. L. V., Oliveira, G., & Behlau, M. (2014). Spectrographic
analysis of the effect of vocal warm-up on the voice of choir girls. Audiology - Communication Research, 19(4), 380–386. doi:10.1590/s2317-64312014000300001372
Gaskill, C. S., & Erickson, M. L. (2008). The Effect of a Voiced Lip Trill on Estimated
Glottal Closed Quotient. Journal of Voice, 22(6), 634–643. doi:10.1016/j.jvoice.2007.03.012
Gish, A., Kunduk, M., Sims, L., & McWhorter, A. J. (2012). Vocal Warm-Up Practices
and Perceptions in Vocalists: A Pilot Survey. Journal of Voice, 26(1), e1–e10. doi:10.1016/j.jvoice.2010.10.005
Guthmiller, J.H. (1986). The goals of vocalization: Developing healthy voices and the
potential for expressive singing. Choral Journal, 26, 13-15. Guzman, M., Laukkanen, A. M., Krupa, P., Horáček, J., Švec, J. G., & Geneid, A.
(2013). Vocal Tract and Glottal Function During and After Vocal Exercising With Resonance Tube and Straw. Journal of Voice, 27(4), 523.e19–523.e34. doi:10.1016/j.jvoice.2013.02.007
Griffin, B., Woo, P., Colton, R., Casper, J., & Brewer, D. (1995). Physiological
characteristics of the supported singing voice: a preliminary study. Journal of Voice, 9, 45-56.
Han, Y., Wang, J., Fischman, D. A., Biller, H. F., & Sanders, I. (1999). Slow tonic muscle
fibres in the thyroarytenoid muscles of human vocal folds; a possible specialization for speech. The Anatomical Record, 256(2), 146–157. doi: 10.1002/(sici)1097-0185(19991001)256:2<146::aid-ar5>3.0.co;2-8
Higgins, M. B., Netsell, R., & Schulte, L. (1998). Vowel-Related Differences in
Laryngeal Articulatory and Phonatory Function. Journal of Speech Language and Hearing Research, 41(4), 712. doi:10.1044/jslhr.4104.712
Hylton, J.B. (1995) Comprehensive Choral Music Education. Englewood Cliffs, NJ:
Prentice Hall.
Univers
ity of
Mala
ya
107
Hoh, J. F. Y. (2005). Laryngeal muscle fibre types. Acta Physiologica Scandinavica, 183(2), 133–149. doi:10.1111/j.1365-201x.2004.01402.x
Horii, Y. (1980). Vocal Shimmer in Sustained Phonation. Journal of Speech Language
and Hearing Research, 23(1), 202.doi:10.1044/jshr.2301.202 Jacobson, B. H., Johnson, A., Grywalski, C., Silbergleit, A., Jacobson, G., Benninger, M.
S., & Newman, C. W. (1997). The Voice Handicap Index (VHI). American Journal of Speech-Language Pathology, 6(3), 66. doi:10.1044/1058-0360.0603.66
James A. Coan, John J.B. Allen (2007). Handbook of Emotion Elicitation and
Assessment. Oxford University Press. Jay Hoffman (2002) Physiological Aspects of Sport Training and Performance. Human
Kinetics. Johnson, Alex & Jacobson, Barbara & Grywalski, Cynthia & Silbergleit, Alice &
Jacobson, Gary & Benninger, Michael. (1997). The Voice Handicap Index (VHI): Development and Validation. American Journal of Speech-Language Pathology, 6. 66-70.
Kapsner-Smith, M. R., Hunter, E. J., Kirkham, K., Cox, K., & Titze, I. R. (2015). A
Randomized Controlled Trial of Two Semi-Occluded Vocal Tract Voice Therapy Protocols. Journal of speech, language, and hearing research : JSLHR, 58(3), 535–549. https://doi.org/10.1044/2015_JSLHR-S-13-0231
Karen Wicklund (2010). Singing Voice Rehabilitation. A Guide for the Voice Teacher
and Speech-Language Pathologist. Cengage Learning. doi:
https://books.google.com.my/books?id=hfD_KBLNKa0C&printsec=frontcover#v=onepage&q&f=false
Karnell MP, Chang A, Smith A & Hoffman H (1997). Impact of signal type of validity
of voice perturbation measures. NCVS Status Progr Rep. 11, 91–94. Karnell MP, Scherer RS, Fischer LB (1991). Comparison of acoustic voice perturbation
measures among three independent voice laboratories. J Speech Hear Res, 34. 781–790.
Univers
ity of
Mala
ya
108
Keith Allan (2003). The Oxford Handbook of the History of Linguistics. OUP Oxford. doi:
https://books.google.com.my/books?id=BQfDosHckzEC&pg=PA129&lpg=PA129&dq=praat+paul+1992&source=bl&ots=2V1oF8q3ly&sig=ACfU3U2I7FYDhlask96v-PgUSNjxW3SE3w&hl=en&sa=X&ved=2ahUKEwjY6tv6rvPnAhXTTX0KHcV8CX0Q6AEwBXoECA4QAQ#v=onepage&q=praat%20paul%201992&f=false
Kersing, W., & Jennekens, F. G. I. (2003). Age-related changes in human thyroarytenoid
muscles: a histological and histochemical study. European Archives of Oto-Rhino-Laryngology, 261(7). doi:10.1007/s00405-003-0702-z
Koufman, J. A., Radomski, T. A., Joharji, G. M., Russell, G. B., & Pillsbury, D. C. (1996).
Laryngeal biomechanics of the singing voice. American Academy of Otolaryngology - Head and Neck Surgery Foundation, Inc., 115, 527-537.
Lathadevi HT, Suresh Pundalikappa Guggarigoudar (2008). Objective Acoustic Analysis
and Comparison of Normal and Abnormal Voices. Journal of Clinical and Diagnostic Research, Vol-12(12): MC01-MC04
Laukkanen, A. M., Horáček, J., & Havlík, R. (2012). Case-study magnetic resonance
imaging and acoustic investigation of the effects of vocal warm-up on two voice professionals. Logopedics Phoniatrics Vocology, 37(2), 75–82. doi:10.3109/14015439.2012.660502
Laukkanen, A. M., Lindholm, P., Vilkman, E., Haataja, K., & Alku, P. (1996). A
physiological and acoustic study on voiced bilabial fricative /β:/ as a vocal exercise. Journal of Voice, 10(1), 67–77. doi:10.1016/s0892-1997 (96)80020-8
Laukkanen, A. M., Horáček, J., Krupa, P., & Švec, J. G. (2012). The effect of phonation
into a straw on the vocal tract adjustments and formant frequencies. A Preliminary MRI study on a single subject completed with acoustic results. Biomedical Signal Processing and Control, 7(1), 50–57. doi: 10.1016/j.bspc.2011.02.004.
Laukkanen, A. M., Titze, I. R., Hoffman, H., & Finnegan, E. (2008). Effects of a
Semioccluded Vocal Tract on Laryngeal Muscle Activity and Glottal Adduction in a Single Female Subject. Folia Phoniatrica et Logopaedica, 60(6), 298–311. doi:10.1159/000170080
Univers
ity of
Mala
ya
109
Lee, S. H., Kwon, H. J., Choi, H. J., Lee, N. H., Lee, S. J., & Jin, S. M. (2008). The singer's formant and speaker's ring resonance: a long-term average spectrum analysis. Clinical and experimental otorhinolaryngology, 1(2), 92–96. https://doi.org/10.3342/ceo.2008.1.2.92
Lundy DS, Roy S, Casiano RR, Xue JW, Evans J. (2000). Acoustic analysis of the singing
and speaking voice in singing students. J Voice. 14:490–493. MacCallum, J. K., Zhang, Y., & Jiang, J. J. (2011). Vowel Selection and Its Effects on
Perturbation and Nonlinear Dynamic Measures. Folia Phoniatrica et Logopaedica, 63(2), 88–97. doi: 10.1159/000319786
Maria Socorro Leonora D. Diaz (2007). Speak Up 6'. Rex Bookstore, Inc. McCrea, C. R., & Morris, R. J. (2007). Effects of Vocal Training and Phonatory Task on
Voice Onset Time. Journal of Voice, 21(1), 54–63. McCrea, C. R., & Watts, C. (2007). Relations of Singing Talent with Voice Onset Time
of Trained and Untrained Female Singers. Perceptual and Motor Skills, 105(1), 133–142.doi:10.2466/pms.105.1.133-142
McCrea, C. R., & Morris, R. J. (2007). Voice onset time for female trained and untrained
singers during speech and singing. Journal of Communication Disorders, 40(5), 418–431. doi:10.1016/j.jcomdis.2006.10.004
McHenry, M., Johnson, J., & Foshea, B. (2009). The Effect of Specific versus Combined
Warm-up Strategies on the Voice. Journal of Voice, 23(5), 572–576. doi:10.1016/j.jvoice.2008.01.003
Mendes, A. P., Brown, W. S., Rothman, H. B., & Sapienza, C. (2004). Effects of singing
training on the speaking voice of voice majors. Journal of Voice, 18(1), 83–89. doi:10.1016/j.jvoice.2003.07.006
Mendes, A. P., Rothman, H. B., Sapienza, C., & Brown, Jr., W. S. (2003). Effects of vocal
training on the acoustic parameters of the singing voice. Journal of Voice, 17, 529-543.
Mezzedimi, C., Spinosi, M. C., Massaro, T., Ferretti, F., & Cambi, J. (2018). Singing
voice: acoustic parameters after vocal warm-up and cool-down. Logopedics Phoniatrics Vocology, 45(2), 57-65. doi:10.1080/14015439.2018.1545865
Univers
ity of
Mala
ya
110
Milbrath, R. L., & Solomon, N. P. (2003). Do Vocal Warm-Up Exercises Alleviate Vocal Fatigue? Journal of Speech Language and Hearing Research, 46(2), 422. doi:10.1044/1092-4388 (2003/035)
Milenkovic, P. (1987). Least Mean Square Measures of Voice Perturbation. Journal of
Speech Language and Hearing Research, 30(4), 529. doi:10.1044/jshr.3004.529 Miller, R. (1990). Warming up the voice. The NATS Journal, 46(5), 22-23. Miller, D. G., Sulter, A. M., Schutte, H. K., & Wolf, R. F. (1997). Comparison of vocal
tract formants in singing and nonperiodic phonation. Journal of Voice, 11(1), 1–11. doi: 10.1016/s0892-1997(97)80018-5
Motel, T., Fisher, K. V., & Leydon, C. (2003). Vocal Warm-up Increases Phonation
Threshold Pressure in Soprano Singers at High Pitch. Journal of Voice, 17(2), 160–167. doi:10.1016/s0892-1997 (03)00003-1
Morrison, M., & Rammage, L. (1994). The management of voice disorders. San Diego,
CA: Singular Publishing Group. 7 (97)80018-5 Muñoz J, Mendoza E, Fresneda MD, Carballo G, Lopez P. (2003). Acoustic and
perceptual indicators of normal and pathological voice. Folia Phoniatr Logop. 55:102–114
Murry, T. (1990). Pitch-matching accuracy in singers and nonsingers. Journal of Voice,
4 (4), 317-321. Naderifar E, Ghorbani A, Moradi N, Ansari H, & Aghadoost O. (2017). Evaluation of
Formant Frequencies in Persian Speaking Children with Different Degrees of Hearing Loss. Shiraz E-Med J. 2017; 18(7):e13094. doi: 10.5812/semj.13094
Narasimhan. S. V & Dr. Nataraja. N. P (2019). Acoustic Analysis of Voice in Subjects
with Hearing Loss. Int J Health Sci Res. 2019; 9(9):194-200. Orlikoff, R. F. (1995). Vocal stability and vocal tract configuration: An acoustic and
electroglottographic investigation. Journal of Voice, 9(2), 173–181. doi: 10.1016/s0892-1997(05)80251-6
Univers
ity of
Mala
ya
111
Omori, K., Kacker, A., Carroll, L. M., Riley, W. D., & Blaugrund, S. M. (1996). Singing power ratio: Quantitative evaluation of singing voice quality. Journal of Voice, 10(3), 228–235. doi: 10.1016/s0892-1997(96)80003-8
Palaparthi, A., Smith, S., Mau, T., & Titze, I. R. (2019). A computational study of depth
of vibration into vocal fold tissues. The Journal of the Acoustical Society of America, 145(2), 881. https://doi.org/10.1121/1.5091099
Paul, B. & Vincent van Heuvan (2001). Speak and unspeak with PRAAT. Glot
International. Poletto, C. J., Verdun, L. P., Strominger, R., & Ludlow, C. L. (2004). Correspondence
between laryngeal vocal fold movement and muscle activity during speech and nonspeech gestures. Journal of applied physiology (Bethesda, Md. : 1985), 97(3), 858–866. https://doi.org/10.1152/japplphysiol.00087.2004
Ramig, L. A., & Ringel, R. L. (1983). Effects of Physiological Aging on Selected
Acoustic Characteristics of Voice. Journal of Speech Language and Hearing Research, 26(1), 22. doi:10.1044/jshr.2601.22
Ribeiro, Frigo, Bastilha & Cielo (2016). Vocal warm-up and cool-down: systematic review. Speech, Language, Hearing Sciences and Education Journal, 8(6):1456- 1465. doi: 10.1590/1982-0216201618617215
Richard Miller (2004). Solutions for Singers. Tools for Performers and Teachers. Oxford
University Press, Inc. Ronald Gillam, Thomas Marquardt & Frederick Martin (2011). Communication Sciences
and Disorders: From Science to Clinical Practice. Jones & Bartlett Learning. Rothenberg. M (1984). Transcripts of the Twelfth Symposium: Care of the Professional
Voice. The Julliard School, New York City, June 6-10, 1983. The Voice Foundation. New York, NY, pp. 15-31, 1984.
Sabol, J., Lee, L., & Stemple, J. (1995). The value of vocal function exercises in the
practice regimen of singers. Journal of Voice, 9, 27-36. Sadhana A.R, Geetha M. (2017) Effect of a Voice Training Program on Acoustic Voice
Parameters in Secondary School Teachers. International Journal of Health Sciences & Research, 7(5):258-263.
Univers
ity of
Mala
ya
112
Sage (2019). The SAGE Encyclopedia of Human Communication Sciences and Disorders. Experimental Research.
doi: http://dx.doi.org/10.4135/9781483380810.n242 Saira, Nazia, Sharmeen and Rukhsana (2017). Acoustic Analysis of Normal Voice
Patterns in Pakistani Adults. Journal of Voice. doi:10.1016/j.jvoice.2017.09.003 Salomoni S, van den Hoorn W, Hodges P (2016) Breathing and Singing: Objective
Characterization of Breathing Patterns in Classical Singers. PLoS ONE 11(5): e0155084. doi: 10.1371/ journal.pone.0155084
Sapir, S. (1989). The intrinsic pitch of vowels: Theoretical, physiological, and clinical
considerations. Journal of Voice, 3(1), 44–51. doi:10.1016/s0892-1997(89)80121-3
Sascha & Pascal (2019). The Oxford Handbook of Voice Perception. Oxford University
Press. Sataloff (2015). Sataloff's Comprehensive Textbook of Otolaryngology: Head & Neck
Surgery: Laryngology. JP Medical Ltd. Sataloff (2017). Professional Voice, Fourth Edition: The Science and Art of Clinical
Care, 3-Volume Set. Plural Publishing. Sataloff (2017). Vocal Health and Pedagogy. Science, Assessment and Treatment. Plural
Publishing. Sataloff (2017). Voice Science, Second Edition. Plural Publishing. Simberg, S., & Laine, A. (2007). The resonance tube method in voice therapy:
Description and practical implementations. Logopedics Phoniatrics Vocology, 32(4), 165–170. doi: 10.1080/14015430701207790
Sorensen, D., & Horii, Y. (1983). Frequency and amplitude perturbation in the voices of
female speakers. Journal of Communication Disorders, 16(1), 57–61. doi: 10.1016/0021-9924(83)90027-8 Stemple, J., Lee, L., Amico, B., & Pickup, B. (1994). Efficacy of vocal function exercises as a method of improving voice production. Journal of Voice, 8, 271-278.
Univers
ity of
Mala
ya
113
Stemple, J. C., Glaze, L., & Klaben, B. (2010). Clinical Voice Pathology: Theory and Management. (4th Ed.). San Diego, CA: Plural Publishing, Inc. Story, B.H. (2003). Using Imaging and Modeling Techniques to Understand the Relation
between Vocal Tract Shape to Acoustic Characteristics. Proceedings of the Stockholm Music Acoustics Conference, August 6-9, 2003 (SMAC 03), Stockholm, Sweden.
Story, B. H., Laukkanen, A. M., & Titze, I. R. (2000). Acoustic impedance of an
artificially lengthened and constricted vocal tract. Journal of Voice, 14(4), 455–469. doi:10.1016/s0892-1997 (00)80003-x
Sundberg, J. (1974). Articulatory interpretation of the “singing formant”. Journal of the
Acoustical Society of America, 55(4), 838–844. doi: https://doi.org/10.1121/1.1914609 Sundberg, J. (1977). The Acoustics of the Singing Voice. Scientific American, 236(3),
82–91. doi: 10.1038/scientificamerican0377-82 Sundberg, J. (1995). The Singer’s Formant Revisited. STL-QPSR 36: 083-096. Sundberg, J. (2001). Level and Center Frequency of the Singer’s Formant. Journal of
Voice, 15(2), 176–186.doi:10.1016/s0892-1997 (01)00019-4 Tay, E., Phyland, D., & Oates, J. (2012). The effect of vocal function exercises on the
voices of ageing community choral singers. Journal of Voice, 26(5), 672.e19- 672.e27.
Teig, E., Dahl, H. A., & Thorkelsen, H. (1978). Actomyosin Atpase Activity of Human
Laryngeal Muscles. Acta Oto-Laryngologica, 85(1-6), 272–281. doi:10.3109/00016487809121450
Teixeira, J. P., & Fernandes, P. O. (2014). Jitter, Shimmer and HNR Classification within
Gender, Tones and Vowels in Healthy Voices. Procedia Technology, 16, 1228–1237. doi:10.1016/j.protcy.2014.10.138
Teixeira, J. P., Oliveira, C., & Lopes, C. (2013). Vocal Acoustic Analysis – Jitter,
Shimmer and HNR Parameters. Procedia Technology, 9, 1112–1122. doi:10.1016/j.protcy.2013.12.124
Univers
ity of
Mala
ya
114
Thomas, L. B., Harrison, A. L., & Stemple, J. C. (2008). Aging Thyroarytenoid and Limb Skeletal Muscle: Lessons in Contrast. Journal of Voice, 22(4), 430–450. doi:10.1016/j.jvoice.2006.11.006
Thorpe, C. W., Cala, S. J., Chapman, J., & Davis, P. J. (2001). Patterns of breath support
in projection of the singing voice. Journal of Voice, 15(1), 86–104. doi: 10.1016/s0892-1997 (01)00009-1 Titze, I. R. (1988). The physics of small‐amplitude oscillation of the vocal folds. The
Journal of the Acoustical Society of America, 83(4), 1536–1552. doi:10.1121/1.395910
Titze, I. R. (1992). Phonation threshold pressure: A missing link in glottal aerodynamics.
The Journal of the Acoustical Society of America, 91(5), 2926–2935. doi:10.1121/1.402928
Titze, I. (1994). Principles of the voice production. Englewood Cliffs, NJ: Prentice Hall. Titze, I. R. (1995). Workshop on acoustic voice analysis: summary statement. Denver,
National Center for Voice and Speech. pp 18– 23. Titze, I. R. (1996). What is vocology? Logopedics Phoniatrics Vocology, 21(1), 5 6. doi:
10.3109/14015439609099196 Titze, I. R. (2001). Acoustic Interpretation of Resonant Voice. Journal of Voice, 15(4),
519–528. doi: 10.1016/s0892-1997(01)00052-2 Titze, I. R. (2004). A theoretical study of f0-f1 interaction with application to resonant
speaking and singing voice. Journal of Voice, 18(3), 292–298. doi:10.1016/j.jvoice.2003.12.010
Titze, I. R. (2006). Voice Training and Therapy with a Semi-Occluded Vocal Tract:
Rationale and Scientific Underpinnings. Journal of Speech Language and Hearing Research, 49(2), 448. doi: 10.1044/1092-4388(2006/035)
Titze, Ingo & Finnegan, Eileen & Laukkanen, Anne-Maria & Jaiswal, Sanyukta. (2002).
Raising lung pressure and pitch in vocal warm-ups: The use of flow-resistant straws. Journal of Singing, 58, 329-338.
Univers
ity of
Mala
ya
115
Titze, I. R., & Laukkanen, A. M. (2007). Can vocal economy in phonation be increased with an artificially lengthened vocal tract? A computer modeling study. Logopedics Phoniatrics Vocology, 32(4), 147–156. doi:10.1080/14015430701439765
Titze, I. R., & Story, B. H. (1997). Acoustic interactions of the voice source with the
lower vocal tract. The Journal of the Acoustical Society of America, 101(4), 2234–2243. doi:10.1121/1.418246
Titze, I. R., & Story, B. H. (2002). Rules for controlling low-dimensional vocal fold
models with muscle activation. The Journal of the Acoustical Society of America, 112(3), 1064–1076. doi:10.1121/1.1496080
Titze IR, Winholtz WS (1993). Effect of microphone type and placement on voice
perturbation measurements. J Speech Hear Res, 36, 1177–1190. Vampola, T., Laukkanen, A. M., Horáček, J., & Švec, J. G. (2011). Vocal tract changes
caused by phonation into a tube: A case study using computer tomography and finite-element modeling. The Journal of the Acoustical Society of America, 129(1), 310–315. doi:10.1121/1.3506347
Van Lierde, K. M., D’haeseleer, E., Baudonck, N., Claeys, S., De Bodt, M., & Behlau,
M. (2011). The Impact of Vocal Warm-Up Exercises on the Objective Vocal Quality in Female Students Training to be Speech Language Pathologists. Journal of Voice, 25(3), e115–e121.doi:10.1016/j.jvoice.2009.11.004
Verdolini, K. (1998). Laryngeal adduction in resonant voice. National Center for Voice
and Speech’s Guide to Vocology. Iowa City: University of Iowa. Vintturi, J., Alku, P., Lauri, E.-R., Sala, E., Sihvo, M., & Vilkman, E. (2001). Objective
Analysis of Vocal Warm-Up with Special Reference to Ergonomic Factors. Journal of Voice, 15(1), 36–53.doi:10.1016/s0892-1997 (01)00005-4
Vurma, A., & Ross, J. (2003). The perception of 'forward' and 'backward placement' of
the singing voice. Logoped Phoniatr Vocol, 28, 19-28. Watson, A. H. D. (2009). The voice: Management and problems. The Biologic of Musical
Performance and Performance-related Injury (Ch. 5). Lanham, MD: Scarecrow Press, 139-192.
Univers
ity of
Mala
ya
116
Watson, P., Hoit, J., Lansing, R., & Hixon, T. (1989). Abdominal muscle activity during classical singing. Journal of Voice, 3, 24-31.
Watts, C., Barnes-Burroughs, K., Estis, J., & Blanton, D. (2006). The singing power ratio
as an objective measure of singing voice quality in untrained talented and nontalented singers. Journal of Voice, 20, 82-88.
Weiss, R., Brown, W. ., & Moris, J. (2001). Singer’s Formant in Sopranos: Fact or
Fiction? Journal of Voice, 15(4), 457–468. doi: 10.1016/s0892-1997(01)00046-7 Welham, N. V., & Maclagan, M. A. (2003). Vocal Fatigue: Current Knowledge and
Future Directions. Journal of Voice, 17(1), 21–30. doi: 10.1016/s0892-1997(03)00033-x
Wolfe, J., Garnier, M., & Smith, J. (2009). Vocal tract resonances in speech, singing and
playing musical instruments. HFSP Journal, 3(1), 6–23. doi: https://doi.org/10.2976/1.2998482 Woods, K., Bishop, P., & Jones, E. (2007). Warm-Up and Stretching in the Prevention of
Muscular Injury. Sports Medicine, 37(12), 1089–1099. doi: 10.2165/00007256-200737120-00006
Yumoto & Gould (1982). Harmonics-to-noise ratio as an index of the degree of
hoarseness. The Journal of the Acoustical Society of America 71(6):1544-9 Zeitels, S. M., Hillman, R. E., Desloge, R., Mauri, M., & Doyle, P. B. (2002).
Phonomicrosurgery in singers and performing artists: Treatment outcomes, management theories, and future directions. Annals of Otology, Rhinology and Laryngology Supplement, 190, 21–40.
Zhang, Z. (2016). Mechanics of human voice production and control. The Journal of the
Acoustical Society of America, 140(4), 2614–2635. doi:10.1121/1.4964509 Ziegler, A., Gillespie, A. I., & Verdolini Abbott, K. (2010). Behavioural Treatment of
Voice Disorders in Teachers. Folia Phoniatrica et Logopaedica, 62(1-2), 9–23. doi:10.1159/000239059
Univers
ity of
Mala
ya